CIVICUS discusses the spread of AI-powered surveillance in Africa with Wairagala Wakabi, executive director of the Collaboration on International ICT Policy for East and Southern Africa (CIPESA) and co-editor of Smart City Surveillance in Africa: Mapping Chinese AI Surveillance Across 11 Countries, the latest report by the African Digital Rights Network (ADRN) and the Institute of Development Studies (IDS).

Wairagala Wakabi

How widespread is AI-powered surveillance in Africa?

Under the guise of reducing crime and fighting terrorism, at least 11 governments have invested over US$2 billion in AI-powered ‘smart city’ surveillance infrastructure: Algeria, Egypt, Kenya, Mauritius, Mozambique, Nigeria, Rwanda, Senegal, Uganda, Zambia and Zimbabwe.

Governments are installing thousands of CCTV cameras linked to central command centres, paired with tools such as automatic number-plate recognition, biometric ID systems and facial recognition to track people and vehicles. The largest known investments are in Nigeria (over US$470 million), Mauritius (US$456 million) and Kenya (US$219 million), though the real total is likely much higher, since surveillance spending is often secret and the report covers only 11 of Africa’s 55 countries.

Despite being presented as tools for crime prevention, counter-terrorism, modernisation and urban management, these are not targeted security measures. They represent a broader shift toward continuous, population-level monitoring of public spaces, rolled out over the past five to ten years almost always without clear legal limits or public debate.

Are these systems achieving their stated purpose?

No, there is no compelling evidence that they have in any of the countries studied. Instead, the data points to a pattern of use that raises serious human rights concerns.

In Uganda and Zimbabwe, AI-powered surveillance including facial recognition is being used to suppress dissent rather than ensure public safety. Activists, critics of the government, opposition leaders and protesters are identified and monitored through this system, even after protests have ended. In Mozambique, smart CCTV systems have reportedly been installed in areas of strong political opposition, suggesting targeted rather than neutral surveillance.

In Senegal and Zambia, countries with relatively low terrorism threats, governments have still invested heavily, which calls into question the stated security rationale.

Across the countries studied, the scale of surveillance far exceeds any actual or perceived security threat, and the infrastructure is consistently being used to monitor dissent and consolidate state control rather than address genuine public safety needs.

Who’s supplying this technology?

While firms from Israel, South Korea and the USA supply surveillance technologies, Chinese companies are the primary suppliers and financiers. They typically offer end-to-end ‘smart city’ packages that include cameras, software platforms, data analytics systems, training and ongoing technical support. Many projects are backed by loans from Chinese state-linked banks, which makes them financially accessible in the short term but creates long-term dependencies on external vendors for maintenance, system management and upgrades.

This model undermines transparency. Procurement processes are opaque and civil society, the public and oversight institutions including parliaments rarely have information about how these systems operate, how data is stored or who has access to it. That lack of accountability is what makes abuse not just possible, but hard to detect or challenge.

What impact is this having on civic space?

This large-scale surveillance of public spaces is not legal, necessary or proportionate to the legitimate aim of providing security. Recording, analysing and retaining facial images of people in public without their consent interferes with their right to privacy and, over time, their willingness to move, assemble and speak freely.

The most immediate consequence is a chilling effect, particularly where civic space is already restricted. Knowing they can be identified and tracked, activists and journalists are less willing to attend protests for fear of later arrest or reprisals, and end up self-censoring. Civil society organisations also report heightened anxiety about the risks for their members and partners.

What should governments and civil society do?

None of the 11 countries studied have a legal framework capable of balancing the state’s security needs with its commitments to protect fundamental human rights. That must change. Governments must adopt clear regulations on surveillance, including restrictions on facial recognition and other AI tools, require independent human rights impact assessments before introducing new systems, make procurement and deployment processes transparent and establish strong oversight mechanisms, including judicial and parliamentary scrutiny, to prevent abuse.

Civil society should continue documenting abuses, raising public awareness and advocating for accountability, while also supporting affected people and communities through digital security support and legal assistance.

Technology-exporting states and donors must enforce stricter controls and safeguards on the export and financing of these tools, support rights-based approaches to digital governance and help fund independent monitoring and advocacy across Africa.

Without urgent action, these systems will continue to expand, and the rights of people across Africa will continue to shrink.

CIVICUS interviews a wide range of civil society activists, experts and leaders to gather diverse perspectives on civil society action and current issues for publication on its CIVICUS Lens platform. The views expressed in interviews are the interviewees’ and do not necessarily reflect those of CIVICUS. Publication does not imply endorsement of interviewees or the organisations they represent.

GET IN TOUCH
CIPESA/Website
CIPESA/Facebook
CIPESA/LinkedIn
CIPESA/Twitter
ADRN/Website
ADRN/Facebook
ADRN/Twitter
IDS/Website
IDS/BlueSky
IDS/Facebook
IDS/Instagram
IDS/LinkedIn

SEE ALSO
Technology: innovation without accountability CIVICUS | State of Civil Society Report 2026
AI governance: the struggle for human rights CIVICUS Lens 11.Sep.2025
Facial recognition: the latest weapon against civil society CIVICUS Lens 23.May.2025

Follow @IPSNewsUNBureau
  

ECA Deputy Executive Secretary for Programme Support, Mama Keita.

By Busani Bafana
TANGIER, Morocco, Apr 1 2026 (IPS)

Africa must move swiftly to harness data and frontier technologies such as Artificial Intelligence (AI) to drive its economic growth and make the continent globally competitive in the digital economy, a senior official at the United Nations Economic Commission for Africa (ECA) has told policymakers.

Opening the Committee of Experts segment of the Conference of African Ministers of Finance, Planning and Economic Development meeting in Tangier, ECA Deputy Executive Secretary for Programme Support Mama Keita emphasised that technological innovation is the key to unlocking Africa’s development potential. Africa has been slow to harness technological innovation to drive industrialisation and economic growth.

“Frontier technologies and innovation are not only useful to unlock Africa’s growth potential and enhance the competitiveness of African economies through productivity growth and diversification,” Keita said. She emphasised that technological innovations can be used to accelerate structural transformation, allowing the much-needed reallocation of resources from low- to high-productivity sectors.

Frontier technologies, including AI, the Internet of Things, and biotechnology, are boosting productivity, enhancing competitiveness, and enabling global economic diversification, but Africa is taking its time to join the party.

Keita, in remarks on behalf of ECA Executive Secretary Claver Gatete, questioned why Africa was not harnessing frontier technologies to utilise its natural resources and tap its youthful population and sizeable markets to boost productivity.

The conference, themed ‘Growth through innovation: harnessing data and frontier technologies for the economic transformation of Africa’, is being held at a critical moment for Africa, which is fast gaining global attention as the next frontier for investment, human capital, and mineral resource development. Despite trade uncertainty, Africa’s economic growth is on the rise.

Keita noted that the conference was an opportunity for policymakers to examine how technology-driven solutions can accelerate structural transformation and deliver more sustainable economic growth in Africa.

Despite averaging 3.5 percent GDP growth between 2000 and 2023, Africa has struggled to convert this expansion into productivity gains. Keita observed that growth has largely been driven by capital and labour accumulation, with little contribution from productivity improvements—an imbalance that innovation and advanced technologies could help correct.

Effective Regulation, Financing and Data Systems Needed

Frontier technologies and data can enable Africa to shift resources from low-productivity sectors to higher-value activities while also improving living standards with effective regulation and financing robust data systems  in place.

Africa suffers from poor data, which constrains effective planning and decision-making for development projects. The ECA’s flagship Economic Report on Africa 2026, to be launched during the conference, argues that harnessing data and technologies like AI, machine learning and robotics is now an imperative for Africa.

Technology Delivers

“There is no doubt that digital platforms, underpinned by frontier technologies such as AI, the Internet of Things, and blockchain, hold significant potential to reduce poverty, generate employment opportunities, promote economic integration, and drive economic growth,” Keita said.

Across the continent, signs are there of how technology innovation is driving development. Digital payment systems and mobile-money platforms are transforming Africa’s economies by lowering transaction costs, boosting efficiency, enhancing access to finance and markets, and advancing financial inclusion.

Nearly 30 per cent of the world’s critical minerals that are essential for clean-energy technologies are in Africa, which gives  the continent a comparative advantage over other continents.

Strategic industries such as digital technologies and telecommunications also depend on the critical minerals, making Africa an indispensable actor in this vital and fast-growing space, she said.

Frontier technologies have boosted crop productivity, enhanced water and land-use efficiency, and promoted climate resilience and adaptation in agriculture.

But Not all is Rosy

Keita said Africa risks falling behind global peers in harnessing the benefits of frontier technologies. AI, for example, is projected to contribute about 5.6 percent to GDP across Africa, Oceania and parts of developing Asia by 2030—lagging behind contributions expected in more advanced economies.

“The adoption of frontier technologies is not all roses, as this is associated with several risks that cannot be ignored,”  Keita warned. “The storage of most of Africa’s data in data centres outside the continent is a big problem, particularly for sensitive data such as medical, financial, and security data, given the sensitivity of such data. It is also costly and results in delays in data transmission.”

Africa currently accounts for less than one percent of global data centre capacity, limiting the deployment of data-intensive technologies like AI, according to the ECA.

“The disruptive effects of new technologies on the African labour market cannot be ignored,” Keita stated, adding that technology tends to cause job losses quickly, while job creation often occurs slowly.

But Africa’s demographic profile of having more young people presents a competitive advantage if it is aligned with the demands of a digital economy.

Globally, AI and automation are expected to create 170 million jobs while displacing 92 million jobs by 2030, resulting in a net gain of 78 million jobs.  Africa can only benefit from these new jobs if it prioritises providing enhanced digital skills training to its population.

&IPS UN Bureau Report

Follow @IPSNewsUNBureau
  

 
CIVICUS discusses China’s tech-enabled repression with Fergus Ryan, a Senior Analyst at the Australian Strategic Policy Institute (ASPI), where he specialises in how the Chinese Communist Party shapes global information environments through censorship, propaganda and platform governance. His research includes a major study on China’s AI ecosystem and its human rights impacts, as well as investigations into China’s use of foreign influencers.

Fergus Ryan

What AI systems is China developing?

Based on our research, China is rapidly developing a multi-layered AI ecosystem designed to expand state control.

Tech giants are building multimodal large language models (LLMs) such as Alibaba’s Qwen and Baidu’s Ernie Bot, which censor and reshape descriptions of politically sensitive images. Hardware companies including Dahua, Hikvision and SenseTime supply the camera networks that feed into these systems.

The state is building what amounts to an AI-driven criminal justice pipeline. This includes City Brain operations centres such as Shanghai’s Pudong district, which process massive surveillance data, as well as the 206 System, developed by iFlyTek, which analyses evidence and recommends criminal sentences. Inside prisons, AI monitors inmates’ facial expressions and tracks their emotions.

AI-enabled satellite surveillance, such as the Xinjiang Jiaotong-01, enables autonomous real-time tracking over politically sensitive regions. Additionally, AI-enabled fishing platforms such as Sea Eagle expand economic extraction in the exclusive economic zones of countries including Mauritania and Vanuatu, displacing artisanal fishing communities.

How does China use AI for censorship and policing?

China relies on a hybrid model of censorship that fuses the speed of AI with human political judgement. The government requires companies to self-censor, creating a commercial market for AI moderation tools. Tech giants such as Baidu and Tencent have industrialised this process: systems automatically scan images, text and videos to detect content deemed to be risky in real time, while human reviewers handle nuanced or coded speech.

In policing, City Brains ingest data from millions of cameras, drones and Internet of Things sensors and use AI to identify suspects, track vehicles and predict unrest before it happens. In Xinjiang, the Integrated Joint Operations Platform aggregates data from cameras, phone scanners and informants to generate risk scores for individuals, enabling pre-emptive detention based on behavioural patterns rather than specific crimes.

On platforms such as Douyin, the state does not just delete content; it algorithmically suppresses dissent while amplifying ‘positive energy’. AI links surveillance data directly to narrative control and police action.

What are the human rights impacts?

These AI systems erode the rights to freedom of expression, privacy and a fair trial.

Historically, online censorship meant deleting a post. Today, generative AI engages in ‘informational gaslighting’. When ASPI researchers showed an Alibaba LLM a photograph of a protest against human rights violations in Xinjiang, the AI described it as ‘individuals in a public setting holding signs with incorrect statements’ based on ‘prejudice and lies’. The technology subtly engineers reality, preventing users accessing objective historical truths.

AI also undermines the right to a fair trial. In courts that lack judicial independence, AI systems that recommend sentences or predict recidivism act as a black box that defence lawyers cannot scrutinise.

Pervasive surveillance changes behaviour even when not actively used, so its chilling effect may be as significant as direct deployment. Knowing their conversations may be monitored, people self-censor online and in private messaging. Emotion recognition in prisons takes this further: people can theoretically be flagged for their internal states of mind. It’s not just actions that are punished, but also thoughts.

Which groups are most affected?

While AI-enabled surveillance affects all people, ethnic minorities such as Koreans, Mongolians, Tibetans and Uyghurs are disproportionately targeted.

Mainstream LLMs are trained primarily in Mandarin, leaving little commercial incentive to develop AI for minority languages. The Chinese state, however, views those languages as a security vulnerability. State-funded institutions, including the National Key Laboratory at Minzu University, are building LLMs in minority languages, not for cultural preservation, but to power public-opinion control and prevention platforms. These scan text, audio and video in Tibetan and Uyghur to detect cultural advocacy, dissent or religious activity.

Feminist activists, human rights lawyers — particularly since the 709 crackdown in 2015 — labour activists and religious minorities including Falun Gong practitioners face disproportionate targeting. Chinese models consistently adopt state-aligned narratives about such groups, labelling Falun Gong a cult and avoiding human rights framing. Since 2020, Hong Kongers have also been subject to National Security Law surveillance using many of the same tools deployed on the mainland, a reminder that this infrastructure can be rapidly extended.

How can activists in China protect themselves?

Protecting oneself inside China is increasingly difficult. AI leaves very few blind spots. But the system is not perfectly omniscient.

Activists have historically relied on coded speech, euphemisms and satire, the classic example being the use of ‘Winnie the Pooh’ to refer to President Xi Jinping. Because AI struggles with cultural nuance and evolving memes, new linguistic workarounds can temporarily bypass automated filters. But this is a relentless game of Whac-a-Mole: Chinese tech companies employ thousands of human content reviewers whose only job is to catch new memes and feed them back into the AI.

The most practical steps are to use VPNs to access blocked platforms, secure communications apps such as Signal and separate devices for sensitive work. None of these are foolproof. VPN use is technically illegal and increasingly detected and Signal can only be accessed via VPN. It helps to keep a minimal digital footprint and communicate face-to-face on sensitive matters. For activists in Xinjiang, however, surveillance is so pervasive that individual precautions offer little protection. Strong international networks and rigorous documentation practices are essential.

Is China exporting these technologies?

China is the world’s largest exporter of AI-powered surveillance technology, marketing these systems globally, particularly to the global south.

The Chinese state is purposefully expanding its minority-language public-opinion monitoring software throughout Belt and Road Initiative countries, effectively extending its censorship apparatus to monitor Tibetan and Uyghur diaspora communities abroad. Chinese companies including Dahua, Hikvision, Huawei and ZTE have deployed surveillance and ‘safe city’ systems across over 100 countries, with Saudi Arabia and the United Arab Emirates among the most significant recipients. Critically, these companies operate under China’s 2017 National Intelligence Law, which requires cooperation with state intelligence, meaning data flowing through these systems could be accessible to Beijing as well as to purchasing governments.

China is also exporting its governance model through the open-source release of its LLMs, embedding Chinese censorship norms into foundational infrastructure used by developers worldwide.

What should the international community do?

The international community must recognise that countering this requires regulatory pushback.

First, democratic states should set minimum transparency standards for public procurement. This means refusing to purchase AI models that conceal political or historical censorship and mandating that providers publish a ‘moderation log’ with refusal reason codes so users know when content is restricted for political reasons.

Second, states should enact ‘safe-harbour laws’ to protect civil society organisations, journalists and researchers who audit AI models for hidden censorship. Currently, doing so can breach corporate terms of service.

Third, strict export controls should block the transfer of repression-enabling technologies to authoritarian regimes, while companies providing public-opinion management services should be excluded from democratic markets. Existing targeted sanctions on companies such as Dahua and Hikvision for their role in Xinjiang should be enforced more rigorously.

Finally, the international community must recognise that Chinese surveillance extends beyond China’s borders. Spyware targeting Tibetan and Uyghur activists in exile is well-documented, as is pressure on family members remaining in China. Rigorous documentation by international civil society remains essential for building the evidentiary record for future accountability.

CIVICUS interviews a wide range of civil society activists, experts and leaders to gather diverse perspectives on civil society action and current issues for publication on its CIVICUS Lens platform. The views expressed in interviews are the interviewees’ and do not necessarily reflect those of CIVICUS. Publication does not imply endorsement of interviewees or the organisations they represent.

GET IN TOUCH
Website
LinkedIn
Twitter/X

SEE ALSO
Technology: innovation without accountability CIVICUS | 2026 State of Civil Society Report
The silencing of Hong Kong CIVICUS Lens 25.Jun.2025
The long reach of authoritarianism CIVICUS Lens 20.Mar.2024

Follow @IPSNewsUNBureau
  

Concept digital technology image with CCTV camera surveillance. Credit: ART STOCK CREATIVE / shutterstock.com Source: Institute of Development Studies, UK

By The Institute of Development Studies
BRIGHTON, UK, Mar 17 2026 (IPS)

A massive expansion of AI-enabled surveillance of public spaces across Africa is violating citizens’ freedoms and the fundamental human right to privacy, warns a new report by the Institute of Development Studies.

African governments are paying billions of dollars to Chinese companies for so-called ‘smart city’ products for public space surveillance – including AI-enabled CCTV and control centres – with at least US$2 billion spent to date by the 11 African countries studied in the report.

The researchers stress that these sophisticated mass surveillance products are being rolled out across Africa without the robust legal frameworks needed to protect human rights. They warn that this lack of protection, coupled with the increased capacity and scale of the smart city mass surveillance leaves government critics, such as the political opposition and independent journalists, at high risk of being tracked and targeted by the state.

The report cites concerns across each of the countries studied. For example, in Zimbabwe, specific groups and government critics fear that facial recognition technologies are used to target them. In Mozambique, research finds that the smart CCTV cameras have been deployed in locations where political opposition is concentrated.

Nigeria is Africa’s largest consumer of Chinese mass surveillance technology, with over US$470 million to date spent on facial recognition and automatic car number plate recognition (ANPR). Mauritius was the second largest buyer out of the 11 African countries studied, spending US$456m, and Kenya the third largest with a spend of US$219m on smart city surveillance technology.

The research reveals that while several countries, including Korea, Israel and the USA, supply public space surveillance technologies, the vast majority of the mass smart city surveillance products used across Africa are supplied and funded by Chinese companies.

Dr Tony Roberts, independent digital rights researcher and co-author of the report, says: “Our new research shows that the rapid growth of smart city surveillance in Africa is occurring without adequate legal regulation or oversight. Unregulated surveillance creates a chilling effect that inhibits the right to peaceful protest and reduces the freedom to speak truth to power and hold governments to account.”

“Digital surveillance of terrorists and the most serious criminals can be justified in the public interest, but installing thousands of smart CCTV cameras for the mass surveillance of all citizens – suspected of no crime – violates important human rights.”

The report details that mass surveillance of public space via smart city technology is being introduced across Africa under the pretence of preventing terrorism or crime, but the researchers found no compelling evidence that the imposing of smart surveillance has led to any reduction in terrorism or serious crime. They also found mass surveillance of public space using smart city technology being used even in countries like Zambia and Senegal that have no terrorist threat or serious crime challenges.

Wairagala Wakabi, Executive Director, CIPESA and co-author of the report, said: “These so-called ‘smart city’ surveillance products are anything but smart for those at risk of being tracked and targeted by them.

“This large scale and invasive AI-enabled surveillance of public spaces is not ‘legal, necessary or proportionate’ to the legitimate aim of providing security. Instead, history shows us that this is the latest tool used by governments to invade the privacy of citizens and stifle freedom of movement and expression.”

“The recording, analysing, and retaining of facial images of individuals in public spaces without their consent interferes with their right to privacy. We need governments to be fully transparent about their procurement and use of smart city technology and ensure that the impacts on human rights have been fully assessed and shared with the public.”

The report was authored by researchers from the African Digital Rights Network and provides the most comprehensive analysis of the use of ‘smart’ city technology in 11 African countries: Algeria, Egypt, Kenya, Mauritius, Mozambique, Nigeria, Rwanda, Senegal, Uganda, Zambia, and Zimbabwe.

Footnote:
At least US$2 billion expenditure on facial recognition and car tracking technologies in 11 countries. The real total is certainly higher because (1) surveillance spending is often secret; (2) no figures were available for two of the 11 countries studied; (3) the public accounts for the other nine countries were incomplete; and (4) this study included only 11 of Africa’s 55 countries.

The Institute of Development Studies (IDS) delivers world-class research, learning and teaching that transforms the knowledge, action and leadership needed for more equitable and sustainable development globally. IDS, in partnership with the University of Sussex, has been named best in the world for Development Studies in the QS World University Rankings by Subject 2025 for the ninth year in a row.

IPS UN Bureau

Follow @IPSNewsUNBureau
  

Millions of children are at risk of facing exploitation and abuse through exposure to and having their images being manipulated through generative AI tools. Credit: Ludovic Toinel/Unsplash

By Oritro Karim
UNITED NATIONS, Feb 10 2026 (IPS)

New findings from the United Nations Children’s Fund (UNICEF) reveal that millions of children are having their images manipulated into sexualized content through the use of generative artificial intelligence (AI), fueling a fast-growing and deeply harmful form of online abuse. The agency warns that without strong regulatory frameworks and meaningful cooperation between governments and tech platforms, this escalating threat could have devastating consequences for the next generation.

A 2025 report from The Childlight Global Child Safety Institute—an independent organization that tracks child sexual exploitation and abuse—shows a staggering rise in technology-facilitated child abuse in recent years, growing from 4,700 cases in the United States in 2023 to over 67,000 in 2024. A significant share of these incidents involved deepfakes: AI-generated images, videos, and audio engineered to appear realistic and often used to create sexualized content. This includes widespread “nudification”, where AI tools strip or alter clothing in photos to produce fabricated nude images.

A joint study from UNICEF, Interpol, and End Child Prostitution in Asian Tourism (ECPAT) International examined the rates of child sexual abuse material (CSAM) circulated online across 11 countries found that at least 1.2 million children had their images manipulated into sexually explicit deepfakes in the past year alone. This means roughly one in every 25 children—or one child in every classroom—has already been victimized by this emerging form of digital abuse.

“When a child’s image or identity is used, that child is directly victimised,” a UNICEF representative said. “Even without an identifiable victim, AI-generated child sexual abuse material normalises the sexual exploitation of children, fuels demand for abusive content and presents significant challenges for law enforcement in identifying and protecting children that need help. Deepfake abuse is abuse, and there is nothing fake about the harm it causes.”

A 2025 survey from National Police Chiefs’ Council (NPCC) studied the public’s attitudes toward deepfake abuse, finding that deepfake abuse had surged by 1,780 percent between 2019 and 2024. In a UK-wide representative survey conducted by Crest Advisory, nearly three in five respondents reported feeling worried about becoming victims of deepfake abuse.

Additionally, 34 percent admitted to creating a sexual or intimate deepfake of someone they knew, while 14 percent had created deepfakes of someone they did not know. The research also found that women and girls are disproportionately targeted, with social media identified as the most common place where these deepfakes are spread.

The study also presented respondents with a scenario in which a person creates an intimate deepfake of their partner, discloses it to them, and later distributes it to others following an argument. Alarmingly, 13 percent of respondents said this behavior should be both morally and legally acceptable, while an additional 9 percent expressed neutrality. NPCC also reported that those who considered this behavior to be acceptable were more likely to be younger men who actively consume pornography and agree with beliefs that would “commonly be regarded as misogynistic”.

“We live in very worrying times, the futures of our daughters (and sons) are at stake if we don’t start to take decisive action in the digital space soon,” award-winning activist and internet personality Cally-Jane Beech told NPCC. “We are looking at a whole generation of kids who grew up with no safeguards, laws or rules in place about this, and now seeing the dark ripple effect of that freedom.”

Deepfake abuse can have severe and lasting psychological and social consequences for children, often triggering intense shame, anxiety, depression, and fear. In a new report, UNICEF notes that a child’s “body, identity, and reputation can be violated remotely, invisibly, and permanently” through deepfake abuse, alongside risks of threats, blackmailing, and extortion from perpetrators. Feelings of violation – paired with the permanence and viral spread of digital content – can leave victims with long-term trauma, mistrust, and disrupted social development.

“Many experience acute distress and fear upon discovering that their image has been manipulated into sexualised content,” Afrooz Kaviani Johnson, a Child Protection Specialist at UNICEF told IPS. “Children report feelings of shame and stigma, compounded by the loss of control over their own identity. These harms are real and lasting: being depicted in sexualised deepfakes can severely impact a child’s wellbeing, erode their trust in digital spaces, and leave them feeling unsafe even in their everyday ‘offline’ lives.”

Cosmas Zavazava, Director of the Telecommunication Development Bureau at the International Telecommunications Union (ITU), added that online abuse can also translate to physical harm.

In a joint statement on Artificial Intelligence and the Rights of the Child, key UN entities, including UNICEF, ITU, the Office of the UN High Commissioner for Human Rights (OHCHR) and the UN Commission of the Rights of the Child (CRC) warned that among children, parents, caregivers and teachers, there was a widespread lack of AI literacy. This refers to the basic ability to understand how AI systems work and how to engage with them critically and effectively. This knowledge gap leaves young people especially vulnerable, making it harder for victims and their support systems to recognize when a child is being targeted, to report abuse, or to access adequate protections and support services.

The UN also emphasized that a substantial share of responsibility lies with tech platforms, noting that most generative AI tools lack meaningful safeguards to prevent digital child exploitation.

“From UNICEF’s perspective, deepfake abuse thrives in part because legal and regulatory frameworks have not kept pace with technology. In many countries, laws do not explicitly recognise AI‑generated sexualised images of children as child sexual abuse material (CSAM),” Johnson said.

UNICEF is urging governments to ensure that definitions of CSAM are updated to include AI-generated content and “explicitly criminalise both its creation and distribution”. According to Johnson, technology companies should be required to adopt what she called “safety-by-design measures” and “child-rights impact assessments”.

She stressed however that while essential, laws and regulations alone would not be enough. “Social norms that tolerate or minimise sexual abuse and exploitation must also change. Protecting children effectively will require not only better laws, but real shifts in attitudes, enforcement, and support for those who are harmed.”

Commercial incentives further compound the problem, with platforms benefitting from increased user engagement, subscriptions, and publicity generated by AI image tools, creating little motivation to adopt stricter protection measures.

As a result, tech companies often introduce guardrails only after major public controversies — long after children have already been affected. One such example is Grok, the AI chatbot for X (formerly Twitter), which was found generating large volumes of nonconsensual, sexualized deepfake images in response to user prompts. Facing widespread, international backlash, X announced in January that Grok’s image generator tool would only be limited to X’s paid subscribers.

Investigations into Grok are ongoing, however. The United Kingdom and the European Union have opened investigations since January, and on February 3, prosecutors in France raided X’s offices as part of its investigation into the platform’s alleged role in circulating CSAM and deepfakes. X’s owner Elon Musk was summoned for questioning.

UN officials have stressed the need for regulatory frameworks that protect children online while still allowing AI systems to grow and generate revenue. “Initially, we got the feeling that they were concerned about stifling innovation, but our message is very clear: with responsible deployment of AI, you can still make a profit, you can still do business, you can still get market share,” said a senior UN official. “The private sector is a partner, but we have to raise a red flag when we see something that is going to lead to unwanted outcomes.”

IPS UN Bureau

Follow @IPSNewsUNBureau
  

IMF Managing Director Kristalina Georgieva at the World Government Summit, Dubai, UAE 3-5 February 2026. Credit: International Monetary Fund (IMF)

By Kristalina Georgieva
DUBAI, United Arab Emirates, Feb 10 2026 (IPS)

It is a pleasure for me to join His Excellency, Minister Al Hussaini in welcoming you to this important dialogue here in the United Arab Emirates—a fast-growing global AI hub. A recent Microsoft study reports that 64 percent of the UAE’s working age population uses AI, which is the highest rate globally.

This illustrates the dynamism we see in the region—and the major investments and partnerships that some of the world’s biggest tech companies are making here.

Why such a huge commitment to this region? Because the UAE and the members of the GCC all understand just how transformative AI can be. They have made systemically significant investments in human capital over the last decades. IMF estimates show that, with the right measures in place, AI could fuel a boost to global productivity of up to 0.8 percentage points per year. This could raise global growth to levels exceeding those of the pre-pandemic period.

Here in the Gulf region, AI could boost non-oil GDP in Gulf countries by up to 2.8 percent. For economies that have long been dependent on hydrocarbon exports, this presents an enormous opportunity to diversify and build new sources of growth.

Now, major technology changes often bring disruption. And sure enough, we can expect disruption from AI. Especially to labor markets. On average, 40 percent of jobs globally will be impacted by AI—either upgraded or eliminated or transformed. For advanced economies, 60 percent of jobs will be affected. This is like a tsunami hitting the labor market.

We are already seeing the evidence: about one in 10 job postings in advanced economies now require at least one new skill. Workers with in-demand skills will likely see productivity and wage gains. This will create more demand for services, and increase employment and wages among low-skilled workers. But middle-skilled jobs will be squeezed.

That means that young people and the middle class will be hit hardest.

We can expect to see a similar divergence between countries. Those with an economic structure conducive to AI adoption—that is, strong digital infrastructure, more skilled labor forces, and robust regulatory frameworks—are likely to experience the largest and fastest benefits. Countries that don’t may get left behind. This is why we gathered here today. AI looks unstoppable.

But whether or not countries can successfully capitalize on AI’s enormous promise is yet to be determined. And this will largely depend on the policy regimes they put in place. So then, what must be done to ensure AI translates into broad-based prosperity for this region?

First, macro policies. Investment and innovation in AI will boost growth. Fiscal policies can support this by strengthening tax systems and by funding research, reskilling, or sector-based training programs. However, tax systems should not encourage automation at the expense of people. Likewise, effective financial regulation will be essential to ensure financial market efficiency and improved risk management.

Second, guardrails. AI needs to be regulated to ensure it’s safe, fair, and trustworthy—but without stifling innovation. Different countries are taking different approaches, ranging from risk-based frameworks to high-level principles. Whatever approach they take, it’s critical that countries coordinate.

That brings me to my third point: cooperation and partnerships. Scale is a big advantage in AI. But you can’t get scale without cooperation among governments, AI researchers and developers, including when it comes to data sharing and knowledge transfer.

Let me conclude. AI will transform our economies. It will present immense opportunities and pose significant risks. And it falls to you, the world’s policymakers, to ensure that the opportunities are maximized for your countries and the risks controlled.

IPS UN Bureau

Follow @IPSNewsUNBureau
  

Artificial intelligence is reshaping how learners, teachers, and creators engage with education across the continent. A new wave of AI innovation transforming learning across countries on the African continent — from chat-based tutors to hybrid hubs and gamified farms. Credit: UNICEF
 
Through initiatives such as Digital Skills for Africa, Lumo Hubs, and Luma Learn, innovators are breaking barriers of access, cost, and language to build inclusive, localized learning systems.

By Franck Kuwonu
UNITED NATIONS, Nov 19 2025 (IPS)

“Sometimes the best way to grasp a concept,” says Chris Folayan, co-founder and executive officer of Luma Learn, “is to learn it in your native language.”

Seventeen-year-old South African Simphiwe is one of more than 10,000 learners already using Luma Learn, an AI-powered tutor platform. For him, artificial intelligence isn’t an abstract idea: it is a personal tutor that is patient, consistent, and always online.

When on his phone, he’s not always chatting with a classmate or scrolling through social media. Many times, he’s studying physics with Luma Learn, that replies instantly, even in IsiZulu, his mother tongue.

Across several countries on the African continent, innovators like Folayan, Nthanda Manduwi, and Anie Akpe are reimagining what education can look like: localised, practical, and accessible to anyone with a phone or connection.

Together, they’re building a new learning ecosystem: one where AI isn’t replacing teachers but multiplying their reach.”

Nthanda Manduwi: Turning digital skills into interactive ecosystems

“I’ve always believed that technology can democratize opportunity,” says Nthanda Manduwi, founder of Digital Skills for Africa (DSA) and Q2 Corporation. “AI gives us a real chance to leapfrog the barriers that have slowed Africa’s progress, from infrastructure gaps to unequal access to training.”

Her journey began with Digital Skills for Africa, a platform designed to equip young people with practical tech competencies from AI and automation to no-code tools and digital marketing.

“Our courses like ‘Effective Use of AI’ or ‘AI and the Future of Digital Marketing’ were created to help learners not only understand AI but actually apply it,” she explains. “You leave with real, marketable skills you can use to build something or get hired.”

But scaling that vision revealed a challenge many edtech startups face. “We realised enthusiasm alone doesn’t pay the bills,” she says. “There was low willingness to pay for courses, even from institutions. So, we had to rethink how to make digital learning sustainable.”

That rethink led to Q2 Corporation, her new venture linking learning with livelihood. Under Q2’s umbrella sits Kwathu Farms—an innovative gamified agricultural simulator where users learn how to manage farms, predict supply chain issues, and test business models before investing real money.

“AI makes the learning immersive,” Ms. Manduwi explains. “Through simulations, learners can see how weather or market shocks affect yield, and how small decisions impact entire value chains. It turns agriculture into a classroom. And a business lab.”

Behind these simulations run Q2’s proprietary engines, NoxTrax and AgroTrax, which apply AI to real-time logistics and resource management. “It’s about showing that AI isn’t just for coders,” she says. “It’s for farmers, small businesses, anyone who wants to think and plan more intelligently.”

Ms. Manduwi’s mission remains rooted in access. “For Africa to truly benefit from AI, it can’t be an elite tool. It must live where people already are: on their phones, in their communities, in local languages.”

Anie Akpe: Creating spaces where AI meets human creativity

Where Ms. Manduwi builds ecosystems, Anie Akpe builds spaces. Through her work with African Women in Technology (AWIT)and Lumo Hubs, Ms. Akpe has spent over a decade helping innovators, especially women, turn curiosity into competence.

“With AWIT, I started by organising conferences across the continent,” she recalls. “We created safe spaces where women could connect with mentors and learn skills that weren’t taught in schools: digital literacy, entrepreneurship, coding, design.”

Soon, even male students began asking to participate. “That’s when I realized it wasn’t just about women in technology. It was about us (Africans) finding a place in a digital world that was changing fast.”

The next step came naturally. “When AI began to disrupt industries, I saw that we couldn’t just talk about skills. We had to create environments where people could use those skills,” she says. “That’s how Lumo Hubs was born.”

Each hub combines education, creativity, and entrepreneurship. “In one space, you might find a student learning AI-assisted graphic design, a seamstress using AI to plan production, and a young podcaster recording a show in a studio powered by the hub,” Ms. Akpe explains. “The model is hybrid, physical and digital, so even small towns can host a Lumo Hub.”

She is also deliberate about sustainability. “Community members pay; students pay less. It’s important that we don’t depend only on grants,” she says. “That balance keeps the hubs alive and the learning continuous.”

At the heart of Lumo Hubs lies mentorship. “You can’t separate technology from human guidance,” Akpe insists. “AI helps scale learning, but mentorship builds confidence.” Her approach remains rooted in empowerment. “AI can level the playing field if used right. A young person in Lagos or Uyo doesn’t have to wait for opportunity. They can create it.”

Chris Folayan: A tutor that never sleeps

For Chris Folayan, the idea behind Luma Learn came from a simple observation: “The continent doesn’t just have an access problem. It has a teaching gap too.”

According to UNESCO, Sub-Saharan Africa will need 15 million new teachers in the next five years to meet demand. “With classrooms that sometimes have over 100 students per teacher, no one can give every child the help they need,” Mr. Folayan says. “That’s where Luma Learn steps in.”

Luma Learn is an AI tutor that runs on WhatsApp, not a separate app.

“We chose WhatsApp for a reason,” he explains. “It’s already on most phones, it’s free to message, works on low bandwidth, and keeps data safe through encryption. That means a child in a rural area can learn without worrying about internet costs or app installations.”

The platform adapts to the learner’s grade level, curriculum, and preferred language. “Whether you need algebra in English or history in Swahili, Luma Learn can teach, quiz, and explain at your level,” he says. “It learns how you learn.”

Mr. Folayan shares two powerful testimonies. In Durban, a mother named Happyness wrote that her son, after years of illness, seizures, and missed schooling, caught up with the rest of the class with help from Luma Learn.

“Every time Vuyo wants to know something about school, we just ask Luma! What’s great is that Luma explains in our native language, IsiZulu.”

In another case, Simphiwe, a Grade 11 student from KwaZulu-Natal, sent over 1,200 messages to Luma. “Luma Learn wasn’t just another study resource,” he said. “It became the personal teaching assistant I desperately needed.”

Shared goals: One vision, many pathways

Three innovators. Three different models. One shared purpose: to make AI work for Africa’s learners, not the other way around. Across their stories, several threads stand out.

First, access—from WhatsApp tutors to open learning hubs to gamified ecosystems that teach real-world problem-solving.

Second, localisation—learning in local languages, within familiar tools, and around community realities.

Third, empowerment—every model links knowledge directly to opportunity.

From Ms. Manduwi’s gamified farms to Ms. Akpe’s creative hubs, to Mr. Folayan’s WhatsApp tutor, future classrooms are already here — decentralised, digital, and deeply human.

As Ms. Manduwi puts it, “We must stop treating AI as something imported. It’s a tool we can mold to fit our own systems.”

Ms. Akpe echoes that sentiment: “Africa doesn’t lack talent. It lacks platforms that meet learners where they are.”

And Mr. Folayan completes the picture: “No teacher wants their student left behind. With AI, we can make sure no one is.”

At the end of the day, a student in Durban learns physics through Luma. A young designer in Uyo experiments with AI tools at a Lumo Hub. A farmer in Lilongwe tests market scenarios on Kwathu Farms. Each represents a different face of the same revolution — a continent using intelligence, both human and artificial, to learn without limits.

As Ms. Akpe says: “The vision is simple: a generation that doesn’t just survive AI disruption but thrives because of it.” And as Ms. Manduwi concludes: “AI is not a threat to Africa. It’s our greatest chance to catch up. And lead.”

Anie Akpe and Chris Folayan were participants at the Global Africa Business Initiative (GABI): Unstoppable Africa2025, held in New York City on the margins of the UN General Assembly in September. The platform helps foster networking, exposure to potential business partners, and garner support for their initiatives.

Source: Africa Renewal, United Nations

IPS UN Bureau

Follow @IPSNewsUNBureau
  

By Deodat Maharaj
GEBZE, Türkiye, Nov 14 2025 (IPS)

Artificial intelligence (AI) is rapidly changing our world. It has helped a few companies in developed countries set record-breaking profits. Last month, Nvidia, a leading US AI company, hit a market value of USD 5 trillion.

Nvidia, together with the other six technology companies known as the Magnificent Seven, reached a market capitalisation of USD22 trillion. This value easily eclipses the combined GDP of the world’s 44 Least Developed Countries (LDCs), Small Island Developing States and Landlocked Developing Countries.

These businesses continue to make massive investments in this transformational technology. Not only are investments being made in AI for the future, but benefits are also already being reaped as it accelerates global commerce and rapidly transforms markets.

According to the World Economic Forum, AI is streamlining supply chains, optimising production, and enabling data-driven trade decisions, giving companies a big competitive edge in global markets.

Thus far, the beneficiaries have been those living in the developed world, and a few developing countries with high technological capacities, like India.

By and large, developing countries have lagged far behind this technological revolution. The world’s 44 LDCs and the Small Island Developing States are those that have been almost completely left out.

According to UNCTAD, LDCs risk being excluded from the economic benefits or the AI revolution. Many LDCs and Small Island Developing States struggle with limited access to digital tools, relying on traditional methods for trade documentation, market analysis, and logistics. This is happening as others race ahead.

This widening gap threatens to marginalize these countries in international trade and underscores the urgency of ensuring they can participate fully in the AI-driven global economy.

AI holds transformative potential for developing countries across sectors critical to economic growth and trade. The World Bank has noted that in agriculture, AI-driven tools can improve crop yields, forecast market demand, and enhance supply chain efficiency. It can also strengthen food security and export earnings. In trade and logistics, AI can optimize operations, reduce transaction costs, and help local producers access new markets.

Beyond commercial applications, AI can bolster disaster preparedness, enabling governments and businesses to allocate resources efficiently and minimize losses. The use of AI can be a game changer in responding to massive natural disasters such as the one caused by Hurricane Melissa in Jamaica a few days ago.

Despite these opportunities, the poorest and most vulnerable countries face significant hurdles in accessing and benefiting from AI. The International Telecommunications Union has noted that many countries lack reliable electricity, broadband connectivity, and computing resources, impeding the deployment of AI technologies. This is compounded by human capacity constraints and limited fiscal space to make the requisite investments.

Given this, what is the best way forward for the world’s poorest and most vulnerable countries? Firstly, policy and governance frameworks for leveraging AI for development transformation are urgently, and we can learn from others.

For example, Rwanda, a leader in the field of using technology to drive transformation has developed a National Artificial Intelligence Policy. Another example is Trinidad and Tobago, which recently established a Ministry of Public Administration and Artificial Intelligence.

Secondly, capacity building, especially for policy leaders, is key. This must be augmented by making the requisite investments in universities and centers of excellence. Given the importance of low-cost and high-impact solutions, building partnerships with institutions in the global south is absolutely vital.

Finally, financing remains key. However, given the downward trends in overseas development assistance, accessing finance, especially grant and concessional resources from other sources will be important. Consequently, international financial institutions, especially the regional development banks, have a critical role to play.

Since the countries themselves are shareholders, every effort should be made to establish special purpose windows of grants and concessional financing to help accelerate adoption of relevant, low-cost, relevant and high-impact AI technological solutions.

In an adverse financing environment, achieving the above will be difficult. This is where Tech Diplomacy comes in and must be a central element of a country’s approach to foreign policy. This will be the subject of another piece.

In summary, AI is shaping and changing the world now. For the poorest and most vulnerable countries, all is not lost. With strategic investments, forward-looking and inclusive policies, and international cooperation via Tech Diplomacy, AI can become a powerful tool for their sustainable growth and development.

Deodat Maharaj, a national of Trinidad and Tobago, is presently the Managing Director of the United Nations Technology Bank for the Least Developed Countries. He can be contacted at: deodat.maharaj@un.org

IPS UN Bureau

Follow @IPSNewsUNBureau
  

Amitabh Behar speaks to IPS at ICSW2025 in Bangkok, Thailand. Credit: Zofeen Ebrahim/IPS

By Zofeen Ebrahim
BANGKOK, Nov 2 2025 (IPS)

Speaking to IPS on the sidelines of the International Civil Society Week in Bangkok (November 1–5), Amitabh Behar, Executive Director of Oxfam International and a passionate human rights advocate, highlighted his concerns about rising inequality, growing authoritarianism, and the misuse of AI and surveillance. Yet, he expressed optimism that, even as civic spaces shrink, young people across Asia are driving meaningful change. He also shared his vision of a just society—one where power is shared, and grassroots movements lead the way.

Excerpts from the interview:

IPS: What does civil society (CS) mean to you personally in today’s global context?

Behar: In an age of grotesque and rising global inequality, civil society is ordinary people challenging elites and the governments that are elected to serve them. It’s the engine that keeps democracy from being just a mere formality that happens at a ballot box every four years.

IPS: What was the role of CS society in the past? How has it evolved? How do you see it in the next decade?

Behar: During Asia’s economic miracle, governments invested in public services while civil society worked alongside unions to defend workers’ rights and speak up for communities. Today, with austerity and rising authoritarianism around the world, civil society is stepping in where governments should be but are currently failing. It runs food banks, builds local support networks, and defends citizens and workers even as basic freedoms and the right to protest are increasingly under attack.

IPS: What do you see as the greatest challenge facing CS today?

Behar: A tiny elite not only controls politics, media, and resources but also dominates decisions in capitals around the world and rigs economic policies in their favor. Rising inequality, debt crises, and climate disasters make survival even harder for ordinary people, while repressive governments actively silence their voices.

IPS: What’s the most significant challenge activists face when it comes to democracy, human rights or inclusion? 

Behar: Authoritarian governments crush dissent and protests with laws, surveillance, and intimidation. AI and digital tools are now being weaponized to track and target and illegally detain protestors, deepen inequality, and accelerate climate breakdown, all while activists risk everything to defend democracy and human rights.

IPS: How can civil society remain resilient in the face of shrinking civic spaces or restrictive laws?

Behar: From protests in Kathmandu to Jakarta, from Dili to Manila, one encouraging theme is emerging: the courage, inspiration, and defiance of young people. Gen Z-led movements, community networks, and grassroots campaigns are winning real change, raising wages, defending workers’ rights, improving services, and forcing action on climate disasters. Despite the immense odds, we will not be silenced. This is our Arab Spring.

IPS: Can you give examples from recent days that indicate that the work of CS is making a difference? Has the outcome been (good or bad) surprising?

Behar: In cities across Asia, Gen Z-led protests are winning higher wages, defending workers’ rights, and forcing local authorities to respond to youth unemployment and climate threats.

IPS: In your experience, what makes partnerships between civil society actors most effective?

Behar: Partnerships work when civil society groups trust each other and put the people most affected at the center. When local networks, youth groups, and volunteers coordinate around community leadership, as in cyclone responses in Bangladesh, for example, decisions are faster, resources reach the right people, and the work actually makes a difference.

IPS: How can civil society collaborate with the government and the private sector without losing its independence?

Behar: Civil society can work with governments and businesses strategically when it genuinely strengthens people’s rights rather than erodes them. But the moment politicians or corporations try to co-opt, stage manage or greenwash their work, civil society can be compromised. Real change only happens when communities set the priorities, not politicians or CEOs.

IPS: What are the biggest strategic choices CSOs need to make now in this shrinking civic space or rising pushback?

Behar: When governments erode rights across the board, from reproductive freedom to climate action, to the right to protest, civil society can’t just stay on the back foot. It must fight strategically, defending civic space, backing grassroots movements, and focusing power, time, and resources where they matter most. The core struggle is inequality, the root of nearly every form of injustice. Striking at it directly is the most strategic way to advance justice across the board.

IPS: In your view, what kinds of alliances (across sectors or geographies) matter most for expanding citizen action in the coming years?

Behar: The alliances that matter are the ones that actually shift power and resources away from the elites. Young people, women, Indigenous communities, and workers linking across countries show governments and corporations they can’t ignore them. When those on the frontlines connect with the wider world, people’s movements stop being small and start changing the rules for everyone.

IPS: How can the marginalized voices be genuinely included in collective action?

Behar: Marginalized voices aren’t there to tick a box or make up the numbers. At spaces like COP in Brazil this year, they should be calling the shots. Indigenous people, women, and frontline communities live through the consequences of rampant inequality every day in every way conceivable. It’s time we pull them up a chair at the table and let them drive the decisions that affect their lives.

IPS: Are emerging technologies or digital tools shaping the work of CS? How? Please mention both opportunities and risks.

Behar: Across Asia, Gen-Z activists are leading protests against inequality and youth unemployment, using digital tools to mobilize, amplify, and organize. But AI and intrusive surveillance now track every post and monitor every march, giving governments even greater powers to violently clamp down on civil society.

IPS: How do you balance optimism and realism when facing today’s social and political challenges?

Behar: I’m optimistic because I see ordinary people, especially young people, refusing to accept injustice. They’re striking, protesting, and building communities that protect each other. But we have to be realistic about the challenge, too. Obscene levels of inequality, worsening climate disasters, and repressive governments make change hard. Yet, time and again, when people rise together, they start to bend the rules in their favor and force the powerful to act.

IPS: What advice would you give to young activists entering this space?

Behar: Keep your fire but pace yourself. Fighting for justice is exhausting, and the challenges can feel endless. Look after your mental health, lean on your community, and celebrate the small wins that can keep you energized for the next challenge. The fight is long, and staying strong, rested, and connected is how you’ll keep on making a difference.

IPS: If you could summarize your vision for a just and inclusive society in one sentence, what would it be?

Behar: A just and inclusive society is one where the powerful can’t rig the rules, the most vulnerable set the agenda, and fairness runs through every policy.

IPS UN Bureau Report

Follow @IPSNewsUNBureau
  

The International Telecommunication Union (ITU) is a specialized agency of the United Nations. Credit: ITU/Rowan Farrell
 
Artificial intelligence holds vast potential but poses grave risks, if left unregulated, UN Secretary-General António Guterres told the Security Council on September 24.

By Chimdi Chukwukere
ABUJA, Nigeria, Oct 14 2025 (IPS)

Recent research from Stanford’s Institute for Human-Centered AI warns that bias in artificial intelligence remains deeply rooted even in models designed to avoid it and can worsen as models grow. From bias in hiring of men over women for leadership roles, to misclassification of darker-skinned individuals as criminals, the stakes are high.

Yet it’s simply not attainable for annual dialogues and multilateral processes as recently provisioned for in Resolution A/RES/79/325 for the UN to keep up to pace with AI technological developments and the cost of this is high.

Hence for accountability purposes and to increase the cost of failure, why not give Tech Companies whose operations are now state-like, participatory roles at the UNGA?

When AI Gets It Wrong: 2024’s Most Telling Cases

In one of the most significant AI discrimination cases moving through the courts, the plaintiff alleges that Workday’s popular artificial intelligence (AI)-based applicant recommendation system violated federal antidiscrimination laws because it had a disparate impact on job applicants based on race, age, and disability.

Judge Rita F. Lin of the US District Court for the Northern District of California ruled in July 2024 that Workday could be an agent of the employers using its tools, which subjects it to liability under federal anti-discrimination laws. This landmark decision means that AI vendors, not just employers, can be held directly responsible for discriminatory outcomes.

In another case, the University of Washington researchers found significant racial, gender, and intersectional bias in how three state-of-the-art large language models ranked resumes. The models favored white-associated names over equally qualified candidates with names associated with other racial groups.

In 2024, a University of Washington study investigated gender and racial bias in resume-screening AI tools. The researchers tested a large language model’s responses to identical resumes, varying only the names to suggest different racial and gender identities.

The financial impact is staggering.

A 2024 DataRobot survey of over 350 companies revealed: 62% lost revenue due to AI systems that made biased decisions, proving that discriminatory AI isn’t just a moral failure—it’s a business disaster. It’s too soon for an innovation to result in such losses.

Time is running out.

A 2024 Stanford analysis of vision-language models found that increasing training data from 400 million to 2 billion images made larger models up to 69% more likely to label Black and Latino men as criminals. In large language models, implicit bias testing showed consistent stereotypes: women were more often linked to humanities over STEM, men were favored for leadership roles, and negative terms were disproportionately associated with Black individuals.

The UN needs to take action now before these predictions turn into reality. And frankly, the UN cannot keep up with the pace of these developments.

What the UN Can—and Must—Do

To prevent AI discrimination, the UN must lead by example and work with governments, tech companies, and civil society to establish global guardrails for ethical AI.

Here’s what that could look like:

Working with Tech Companies: Technology companies have become the new states and should be treated as such. They should be invited to the UN table and granted participatory privileges that both ensure and enforce accountability.

This would help guarantee that the pace of technological development—and its impacts—is self-reported before UN-appointed Scientific Panels reconvene. As many experts have noted, the intervals between these annual convenings are already long enough for major innovations to slip past oversight.

Developing Clear Guidelines: The UN should push for global standards on ethical AI, building on UNESCO’s Recommendation and OHCHR’s findings. These should include rules for inclusive data collection, transparency, and human oversight.

Promoting Inclusive Participation: The people building and regulating AI must reflect the diversity of the world. The UN should set up a Global South AI Equity Fund to provide resources for local experts to review and assess tools such as LinkedIn’s NFC passport verification.

Working with Africa’s Smart Africa Alliance, the goal would be to create standards together that make sure AI is designed to benefit communities that have been hit hardest by biased systems. This means including voices from the Global South, women, people of color, and other underrepresented groups in AI policy conversations.

Requiring Human Rights Impact Assessments: Just like we assess the environmental impact of new projects, we should assess the human rights impact of new AI systems—before they are rolled out.

Holding Developers Accountable: When AI systems cause harm, there must be accountability. This includes legal remedies for those who are unfairly treated by AI. The UN should create an AI Accountability Tribunal within the Office of the High Commissioner for Human Rights to look into cases where AI systems cause discrimination.

This tribunal should have the authority to issue penalties, such as suspending UN partnerships with companies that violate these standards, including cases like Workday.

Support Digital Literacy and Rights Education: Policy makers and citizens need to understand how AI works and how it might impact their rights. The UN can help promote digital literacy globally so that people can push back against unfair systems.

Lastly, there has to be Mandates for intersectional or Multiple Discriminations Audits: AI systems should be required to go through intersectional audits that check for combined biases, such as those linked to race, disability, and gender. The UN should also provide funding to organizations to create open-source audit tools that can be used worldwide.

The Road Ahead

AI is not inherently good or bad. It is a tool, and like any tool, its impact depends on how we use it. If we are not careful, AI could lengthen problem-solving time, deepen existing inequalities, and create new forms of discrimination that are harder to detect and harder to fix.

But if we take action now—if we put human rights at the center of AI development—we can build systems that uplift, rather than exclude.

The UN General Assembly meetings may have concluded for this year, the era of ethical AI has not. The United Nations remains the organization with the credibility, the platform, and the moral duty to lead this charge. The future of AI—and the future of human dignity—may depend on it.

Chimdi Chukwukere is an advocate for digital justice. His work explores the intersection of technology, governance, Big Tech, sovereignty and social justice. He holds a Masters in Diplomacy and International Relations from Seton Hall University and has been published at Inter Press Service, Politics Today, International Policy Digest, and the Diplomatic Envoy.

IPS UN Bureau

Follow @IPSNewsUNBureau
  

Least Developed Countries account for less than 1 percent of world trade. Credit: Ali Mkumbwa/Unsplash

By Deodat Maharaj
GEBZE, Türkiye, Aug 22 2025 (IPS)

Artificial intelligence and the use of frontier technologies are already transforming trade and boosting prosperity, particularly for developed and some developing countries. This ranges from the digital exchange of documents, the digitalisation of trade processes and leveraging online platforms to fast-track cross-border trade.

The rapid adoption of new technologies will further consolidate the dominance of world trade by developed economies, which currently account for roughly 74 percent of global trade, according to the United Nations Conference on Trade and Development (UNCTAD). The world’s 44 Least Developed Countries (LDCs), with a population of an estimated 1.4 billion people, are seeing a different trajectory altogether. According to the World Trade Organisation, they account for less than 1 percent of the world’s merchandise trade. LDCs continue to reel from the relentless onslaught of bad news, including increased protectionist barriers.

Deodat Maharaj, Managing Director of the United Nations Technology Bank for the Least Developed Countries.

UNCTAD has estimated that tariffs on LDCs will have a devastating consequence, possibly leading to an estimated 54 percent reduction in the exports from the world’s poorest countries.

In this dire situation, exacerbated by declining overseas development assistance, what does an LDC do to survive in this difficult trade environment?

To start with, they must continue to advocate globally for fairer terms of trade. At the same time, they need to be more aggressive in addressing matters for which they have control. Otherwise, the status quo will leave their people in a perpetually disadvantageous situation. Imagine paying three times more than your competitors just to ship a single crate of goods across a border. For millions of entrepreneurs in the world’s LDCs, it is the everyday cost of doing business. Technology offers a way out in reducing these high costs.

Indeed, when the international community gathered in Sevilla for the Fourth International Conference on Financing for Development (FfD4) in July 2025, one truth stood out: Technology is no longer a luxury—it is a prerequisite for effective participation in global trade. The outcome document was clear that for the world’s 44 LDCs, bridging infrastructure gaps, building domestic technological capacity, and leveraging science, technology, and innovation are vital to unlocking trade opportunities.

So, given the challenges and opportunities, what forms the core elements of an action agenda for LDCs to leverage trade to generate jobs and opportunities for their people?

Firstly, there is a need to pivot to digital solutions, which can dramatically reduce trade costs and open new markets. According to the World Bank, paperless customs and single-window systems have been proven to cut clearance times by up to 50 percent, reducing bureaucracy that stifles commerce. In Benin, automating port procedures reduced processing time from 18 days to just three days (World Bank). E-commerce platforms, when paired with secure payment systems and targeted training, have shown remarkable potential.

Secondly, invest in digital infrastructure. The data suggest that LDCs still have a lot of catching up to do. The solution is for development partners and the international financial institutions to steer more resources in this area with a fixed percentage of resources, say, 15 percent of a country’s portfolio dedicated to boosting digital infrastructure.

Thirdly, focus on value addition and reduce transition away from the export of raw commodities. This in turn requires the human resource capacity to spur innovation and creativity. Boosting investment in research and development can pay rich dividends.

According to the World Economic Forum, LDCs invest less than 1 percent of GDP in research and development compared to developed countries. The Republic of Korea invests 4%.

Finally, for LDCs to enter the technological age, their businesses must lead the way. It is difficult to do so in some countries like Burundi, where internet penetration is a mere 5 percent of the population. The average internet penetration is around 38 percent. So, in addition to digital infrastructure, support must be provided to micro-, small and medium-scale enterprises to benefit from the opportunities provided by technology to boost trade, thereby creating jobs and opportunities. This includes the establishment of incubators to support this business sector, boosting their technological capacities to trade and profile their businesses on digital platforms, and helping them to deliver services created by the digital economy. Rwanda has been a pioneer in this regard.

Of course, technology alone will not address all the challenges faced by LDCs. However, by delivering cost-efficient solutions, it can help level the playing field and drive transformation. It is time for the international community and development partners to back their words with action in helping LDCs advance this agenda. Since LDCs represent an emerging market of 1.4 billion people, when they rise, everyone else will rise with them.

Deodat Maharaj, a national of Trinidad and Tobago is the Managing Director of the United Nations Technology Bank for the Least Developed Countries and can be reached at: deodat.maharaj@un.org

Follow @IPSNewsUNBureau
  

UN Secretary General António Guterres addresses the session “Strengthening multilateralism, economic - financial affairs and artificial intelligence” on July 6 at the 17th summit of BRICS in Rio de Janeiro. For the first time ever, artificial intelligence was a major topic of concern at the BRICS summit. Credit: UN Photo/Ana Carolina Fernandes

By Naomi Myint Breuer
UNITED NATIONS, Jul 14 2025 (IPS)

Artificial intelligence (AI) is rapidly developing and leaving its mark across the globe. Yet the implementation of AI risks widening the gap between the Global North and South.

It is projected that the AI market’s global revenue will increase by 19.6 percent each year. By 2030, AI could contribute USD 15.7 trillion to the global economy. However, the increases to nations’ GDP will be unequally dispersed, with North America and China experiencing the most gains while the Global South gains far less.

The risks of AI to the Global South

Due to smaller capacities to fund research, development and implementation, fewer countries in the Global South are adopting AI technology. Access to affordable AI compute to train AI models is one of the AI field’s greatest barriers to entry in the Global South, according to the 2024 UN report, “Governing AI for Humanity.”

Further, AI is designed to create profitable market extraction that does not benefit the global majority, according to Vilas Dhar, President and Trustee of the Patrick J. McGovern Foundation. As countries in the Global North are AI’s primary investors, it is being developed to address their needs.

“The result is a quiet erosion of political and economic autonomy,” he said. “Without deliberate intervention, AI risks becoming a mechanism for reinforcing historical patterns of exploitation through technical means. It also risks losing the incredible value of diverse, globally minded inputs into designing our collective AI future.”

Across the world, people risk losing their jobs to AI, but many countries in the Global South are reliant on labor intensive industries, and AI poses a greater threat to increasing unemployment and poverty. Particularly children, women, youths, people with disabilities, older workers, creatives and people with jobs susceptible to automation are at risk.

According to Daron Acemoglu, professor at the Massachusetts Institute of Technology, labor-replacing AI poses a greater threat to workers in the developing world, as capital-intensive technology may not be useful in these nations where oftentimes capital is scarce and labor is abundant and cheap. Technology that prioritizes labor-intensive production is better suited to their comparative advantage.

“Because advanced economies have no reason to invest in such labor-intensive technologies, the trajectory of technological change will increasingly disfavor poor countries,” he said.

If these trends continue, these nations will experience increased unemployment and fall behind in the deployment of capital-intensive AI, due to limited financial resources and digital skill sets. More AI policies and guidelines, as well as education on data privacy and algorithmic bias, could assist in reducing this inequality.

Evidently, AI threatens to widen the gap between the Global North and South, as AI capacities are consolidated within a small group of institutions and regions. In Dhar’s view, AI will need to be designed to serve people and problems rather than be focused on profit maximization.

“If left unaddressed, this imbalance will cement a way of thinking about the world that mirrors the development of the Internet or social media – a process we do not want to replicate,” Dhar said.

Opportunities of the new technology

But the development of AI also poses opportunities for the Global South.

AI could design context-specific systems for local areas in the Global South that are not just based on the Global North, according to Dhar. “It can unlock new models of inclusion and resilience,” he said.

For example, AI could aid farmers in decision-making by informing them of weather and drought predictions using geospatial intelligence, as well as of marketing price information. AI could also help train farmers and other producers. It can also be used to improve education and healthcare in nations where these are major issues harming their populations and stunting development.

Acemoglu said that AI should be developed to complement rather than replace human labor for these benefits to become possible. “That will require forward-looking leadership on the part of policymakers,” he said.

AI in conflict

AI is also starting to make an appearance in conflict. In Ukraine, autonomous drones are being used, which are capable of tracking and engaging enemies, as well as BAD.2 model robot dogs, which are ground drones that can survey areas for enemies. Autonomous machine guns are also used, in which AI helps spot and target enemies.

The use of AI in conflict poses an ethical dilemma. AI could protect human lives on one side of the conflict but pose a great threat to the lives on the other end of the battlefield. This also raises the question of whether AI should be given the power to engage in harm.

But perhaps the use of AI can reduce the number of people engaging in conflicts harming developing countries and move these people to other sectors where they can realize more potential and aid their country’s economic development.

What international frameworks should do

Clear international frameworks must be established to prevent a rise in inequality and a greater gap between the Global North and South.

For the first time ever, AI was a major topic of discussion at the 17th BRICS summit, which serves as a coordination forum for nations from the Global South, in Rio de Janeiro. BRICS member countries signed the Leaders’ Declaration on Global Governance of Artificial Intelligence, which presents guidelines to ensure AI is developed and used responsibly to advance sustainability and inclusive growth.

The declaration called on members of the UN to promote including emerging markets and developing countries (EMDCs) and the Global South in decision-making regarding AI.

“New technologies must operate under a governance model that is fair, inclusive, and equitable. The development of AI must not become a privilege for a handful of countries, nor a tool of manipulation in the hands of millionaires,” Brazilian president Luiz Inácio Lula da Silva said at the summit.

However, the UN report “Governing AI for Humanity” found that 118 countries, most of which are in the Global South, were not part of a sample of non-UN AI governance initiatives, while seven countries, all of which are in the Global North, were included in all initiatives.

According to Dhar, global governance must create a more equitable distribution of power that entails sharing ownership and embedding the Global South at every level of institutions, agreements and investments, rather than simply for consultation. These nations must also be aided in building capacity, sharing infrastructure, scientific discovery and participation in creating global frameworks, he said.

In his remarks at the BRICS summit, UN Secretary-General António Guterres expressed his concern over the weaponization of AI and stressed the importance of AI governance that is focused on equity. He said in order for this to be done, the current “multipolar world” must be addressed.

“We cannot govern AI effectively—and fairly—without confronting deeper, structural imbalances in our global system,” Guterres said.

Dhar emphasized that the inclusion of every person in the conversation on AI is crucial to creating legitimate global technological governance.

The future of AI is being negotiated with immediacy and urgency,” Dhar said. “Whether it becomes a force for collective progress or a new vector for inequality depends on who is empowered to shape it.”

IPS UN Bureau Report

Follow @IPSNewsUNBureau
  

A digital meeting by Brazil’s Ministry of Science and Technology to discuss the use of artificial intelligence in the public sector. Remote work and debates have also increased the demand for digital infrastructure by boosting long-distance communication. Credit: Rodrigo Cabral / Ascom MCTI

By Mario Osava
RIO DE JANEIRO, May 30 2025 (IPS)

Brazil hopes to soon reap benefits of its largely renewable energy matrix. Data centers, whose demand is growing with the strides made by artificial intelligence, are the new frontier for these still-uncertain investments."The most serious issue in the government's program is that it aims to subsidize data centers for big tech companies... they propose bringing in data centers for Google, Facebook, Microsoft, and others, with all the benefits." — Carlos Afonso.

This is even a matter of “digital sovereignty,” not just for Brazil, according to Dora Kaufman, a professor in the program on intelligent technologies and digital design at the Pontifical Catholic University of Sao Paulo.

Nearly 60% of all Brazilian data processing currently takes place in the United States—and the figure continues to rise—posing a serious risk, as a natural disaster or government blockade could paralyze the country, she warned. “The probability of it happening is low, but the impact would be huge,” she told IPS by phone from São Paulo.

The National Data Center Policy is expected to change this scenario, according to the Brazilian government, which has promised to soon unveil the program. Its potential could attract two trillion reais (around US$350 billion) over the next 10 years, claims Finance Minister Fernando Haddad.

Exemptions from federal taxes and reduced import duties on equipment are among the incentives the government will offer investors. These measures anticipate policies already outlined in the recently approved tax reform, which will fully take effect by 2033.

The abundance of renewable energy, water, and land could also serve as a major draw in a world increasingly demanding sustainability in new projects.

Engineering and computer science students in Rio de Janeiro will form an essential workforce for the expanding digital economy, fueled by the government’s policy to encourage the proliferation of data centers in Brazil. Credit: Tomaz Silva / Agência Brasil

High Costs in Brazil 

Processing data in Brazil is 25% more expensive than abroad, primarily due to the tax burden, noted Kaufman. Removing this obstacle would pave the way for a surge in data centers, as “we have more than enough renewable energy and water,” she argued.

“Brazil has everything it takes to host many data centers, and the challenges are solvable. We need them not just to develop artificial intelligence but also for the growing digitalization of government and businesses,” she emphasized.

However, the voracious energy and water demands of digital infrastructure—especially for AI—are raising concerns among environmentalists and experts in energy and communications.

“Brazil first needs to implement a real energy transition. So far, we’ve only added renewable sources alongside fossil fuels. A just transition remains a huge challenge, requiring the electrification of transport—a priority due to the climate crisis,” said Alexandre Costa, a professor at the Federal University of Ceará in northeastern Brazil.

TikTok plans to set up a data center in Caucaia, a city of 355,000 residents in Ceará. Just 35 kilometers away, the Pecém port—which includes an industrial zone—has plans for a green hydrogen production hub, another major consumer of water and electricity.

Pecém already hosts a thermoelectric plant and a steel mill, both of which are highly water-intensive.

In the industrial zone of the Pecém port, in Ceará, wind turbine blades are manufactured. Nearby, there are plans to produce green hydrogen for export to Europe. The high consumption of electricity and water worries environmentalists in this and other regions of Brazil where large data centers are planned. Credit: Mario Osava / IPS

 Fossil Fuels Still Dominate

The Northeast, Brazil’s poorest region, has become an attractive location for projects claiming to be sustainable, as it is already the country’s largest wind power producer and holds vast potential for solar energy.

However, the exploitation of strong, steady winds and abundant sunlight has already sparked criticism and protests from local communities. The expansion of these projects is encroaching on increasing amounts of land, creating conflicts with local populations and small-scale farming, noted Costa, a physicist specializing in meteorology and climate change.

Nationally, renewable sources accounted for 86.1% of electricity consumption in 2022, according to the government’s Energy Research Company. However, fossil fuels still made up 52.7% of Brazil’s total energy matrix, dominated by oil and natural gas, while coal held a small 4.4% share.

This means Brazil, where freight transport is still heavily reliant on diesel trucks, still has a long way to go in reducing fossil fuel consumption. This transition will require even more electricity.

Data centers will bring additional energy demand to an economy already anticipating a surge in consumption—driven by green hydrogen projects, artificial intelligence, and vehicle electrification, Costa warned IPS in a phone interview from Fortaleza, Ceará’s capital.

The same applies to water resources. “There’s no way to meet an infinite demand for these inputs,” he stressed. In his view, Brazil lacks an energy model that balances new demands, priorities, and the need for an increasingly clean energy matrix.

The electrification of vehicles is increasing electricity demand. Data centers create additional pressure on power generation from renewable sources to meet Brazil’s goals of reducing greenhouse gas emissions. Credit: Marcelo Camargo / Agência Brasil

Dependence 

“The most serious issue in the government’s program is that it aims to subsidize data centers for Big Techs. We need them for our national networks, yet they’re proposing to bring in data centers for Google, Facebook, Microsoft, etc., with all the benefits,” criticized Carlos Afonso, a communications technology expert and one of the pioneers of the internet in Brazil.

He pointed to the lack of such infrastructure for public entities like Serpro (Data Processing Service) and Dataprev (social security database), which are vital for government operations, as well as the National Research Network that connects universities and other scientific and innovation institutions.

“Will they have to rely on data centers from these Big Techs in Brazil?” he questioned in a conversation with IPS.

It appears that both the government’s program for this sector and its green hydrogen initiative are primarily designed to meet external demands, with the goal of creating exportable goods and services.

This is why Kaufman argues for imposing conditions on data centers established in Brazil, such as sustainability based on renewable energy and zero greenhouse gas emissions, energy efficiency, and  allocating at least 10% of installed capacity to the domestic market.

The expert believes that the large data centers to be installed in Brazil will primarily serve AI training, which minimizes latency, the milliseconds of delay in long-distance communication from origin to destination.

But the reality—both in Brazil and globally—in the digital economy is one of deep dependence on the United States, a situation exacerbated by the policies of President Donald Trump, who prioritized the interests of the United States above all else, even international treaties.

“Three Big Tech companies from the United States—AWS/Amazon, Microsoft, and Google—control 63% of global data processing, forming a true oligopoly,” emphasized Kaufman. That dominance is expected to grow to 80%, she added.

According to the global statistics portal Statista, as of March 2025, the United States had 5,426 data centers—more than 10 times the number in Germany (529), the UK (523), or China (449).

The imbalance is even starker in hyperscale data centers, those occupying more than 930 square meters and housing over 5,000 servers. By the end of 2024, the United States accounted for 54% of global processing capacity, compared to 16% for China and 15% for Europe, according to Synergy Research Group.

In 2024 alone, 137 new data centers were built—a 13.7% growth rate—in a trend expected to continue, driven largely by advancements in artificial intelligence, notes the analytics and consulting firm based in the United States.

The infrastructure powering the digital economy, already connecting two-thirds of humanity and expanding rapidly with innovations like cloud computing and AI, remains largely unseen.

While cables, including intercontinental submarine lines, satellites, and telecom networks are well-known, data centers—the “brains” that store, process, and distribute information—operate in relative obscurity. Yet, they have become massive and strategically critical as global data traffic surges exponentially.

Roseau, the capital of Dominica in the Eastern Caribbean. The UNDP Human Development Report 2025 shows that countries in Latin America and the Caribbean have made progress but still face challenges like inequality and slow growth, with AI considered a key opportunity to accelerate inclusive development. Credit: Alison Kentish/IPS

By Alison Kentish
DOMINICA, May 22 2025 (IPS)

The United Nations Development Programme’s 2025 Human Development Report (HDR) says crises such as the Covid-19 pandemic have contributed to ‘the flatlining of decades of progress in the Human Development Index,’ with Latin America and the Caribbean facing unique challenges and opportunities.

Despite these challenges, the document, titled “A matter of choice: People and Possibilities in the age of AI,” states that artificial intelligence (AI) is a powerful tool to improve lives and close persistent gaps.

Lead author Pedro Conceiçāo described a ‘triple development squeeze’ affecting many countries.

“Difficulties accessing external financing, shrinking job creation opportunities and increased trade volatility,” he explained. “The opportunities of many countries to export to international markets, which is an important driver of development or has historically been, those opportunities are also narrow.”

Amid these pressures, AI emerges as a double-edged sword. According to a recent UNDP survey, “Up to two-thirds of people in low, medium, and high HDI countries expect artificial intelligence to become an important part of their lives within the next year—in health, education, and standard of living,” Conceição noted. He said the report and survey emphasize that “what matters less is the technology and more the choices that are made to ensure that AI advances human development.”

The report’s recommendations are clear:

• Build a complementary economy where AI empowers, not replaces, people.
• Drive innovation with intent, using AI to boost creativity and scientific progress.
• Invest in digital capabilities so everyone can thrive in an AI-driven world.

The Latin America and the Caribbean Situation

UNDP Regional Director for Latin America and the Caribbean Michelle Muschett outlined the region’s progress and pressure points.

“Latin America and the Caribbean consolidated its second year of recovery after the pandemic, moving from 0.783 in 2022 to 0.8 in the Human Development Index regionally in 2023,” she said. However, she cautioned, “Progress continues, but it remains slower than before the pandemic.”

The region stands out for its high human development scores—19 countries are classified as high, and 10 as very high. But Muschett warns, “Both development and democracy are under probably unprecedented pressure in the history of development of our region.”

She said this should serve as both a warning and a call to action.

“It’s a clear call to thinking and rethinking those institutions, public policies, processes, and the tools we have so that that pressure can become a positive force that moves us along the line of progress and shared prosperity.”

Muschett is candid about the region’s digital disparities. “We see already today the deep difference in terms of coverage when we compare rural areas with urban areas in Latin America and the Caribbean,” she says. “The highest quintile in terms of income has more than twice the access to AI than the lowest quintile. So we have a warning signal that is very important.”

To address digital gaps, the report calls for closing connectivity gaps, especially in rural and low-income areas; investment in digital literacy and lifelong learning; and ensuring that data is reliable and free from bias through strong, inclusive governance frameworks.

“This has to be a central priority of public policies,” Muschett urges. “Strategic alliances with other sectors of society—academia, private sector—become absolutely essential.”

A Resilient Future

Muschett says the UNDP is preparing to launch an “atlas of AI focused on human development,” offering policymakers tools to make informed, inclusive choices.

The message is clear: While the region faces significant challenges, deliberate action can shift the view of AI as a pressure point into a powerful driver of progress.

“The difference between one and the other is precisely in the deliberate decision we make as a region… whether it’s a huge threat or an unprecedented opportunity,” she said.

The message is clear: by fostering innovation, empowering individuals, and putting inclusion at the forefront, Latin America and the Caribbean have the potential to transform current obstacles into future possibilities—and become a worldwide model for leveraging technology to benefit all.

IPS UN Bureau Report

Follow @IPSNewsUNBureau
  

Excerpt:

Credit: Cynthia R Matonhodze/IMF Photo

By Thomas Cantens and Herve Tourpe
WASHINGTON DC, Feb 26 2025 (IPS)

New technologies have the potential to improve the relationship between governments and citizens. Tax portals, customs IT systems and online services have simplified interactions with public authorities, reduced bureaucratic hurdles, and increased transparency. Now, generative artificial intelligence (GenAI) is emerging as the next transformative force.

Known for its ability to understand and produce human language, GenAI opens possibilities that go beyond simple automation. However, in an area as politically sensitive as taxation, it also raises important questions that could quickly undermine trust.

Tax authorities are beginning to explore GenAI, though most efforts are still at an early, experimental stage. The most evident area so far has been on improving communication with taxpayers.

In Singapore, a virtual assistant answers tax questions in multiple languages and has cut call-center inquiries by half. Korea has deployed an AI guide to help citizens file and pay taxes. In France, AI can analyze incoming emails and propose draft responses for civil servants to validate.

While these applications are promising, a more profound question emerges: Can GenAI significantly alter the relationship between governments and citizens? Furthermore, how will it influence the way citizens experience and perceive taxation—a politically sensitive process that is governed by law yet deeply intertwined with social norms and practices?

What’s new with GenAI?

Most AI systems currently used by tax and customs authorities are predictive and built for a single function. They analyze large sets of structured data—like past tax declarations or transactions—to produce things like risk scores to indicate possible fraud.

By contrast, GenAI is a generalist system that understands almost all forms of information and is designed to interact with humans in any language. It can handle a range of tasks, from drafting letters to providing interactive guidance about tax regulations and assisting officers in their investigations.

By training a GenAI agent with legal texts, tax codes, operating procedures, and internal guidelines, administrations can adapt it to specific needs. The result is a dynamic system capable of understanding and producing content that both civil servants and taxpayers can interact with.

Transforming the State-Society Relationship

While AI tools already in use often enhance efficiency, they have not fundamentally changed the way revenue authorities work or engage with citizens. They mostly replaced manual tasks or systems for econometric or statistical modelling.

With GenAI, there are more profound implications. Internally, it can help tax and customs officials to focus on analytical and judgment-based roles, allowing them to become oversight specialists and increasing their productivity.

Externally, it can reduce the knowledge gap between administrations and taxpayers, aiding in the interpretation of complex provisions, navigating laws, identifying deductions, and even auto-filling forms.

For low-income countries, GenAI offers the opportunity to drive organizational reforms and leapfrog into the most modern systems. For example, in Madagascar, the customs authority wants to use GenAI to improve risk management, combat fraud and increase revenue, using data accumulated over 10 years to train its system.

The human-like interactions offered by AI chat tools can personalize the process, as shown in Singapore and Korea, where users can ask questions and receive plain language replies. Citizens’ organizations, academics, and political parties can also use GenAI to examine proposed reforms, compare scenarios, and engage in deeper policy debates.

This two-way transformation could increase overall trust, making taxation feel less like a frustrating obligation and more like a shared responsibility of both taxpayers and governments.

Preconditions for success

Despite its potential, GenAI also comes with challenges. Issues related to data quality, ethics, privacy concerns and hallucinations (i.e., incorrect results) must be addressed to reinforce and not erode trust. For instance, Korea’s approach—directing particularly sensitive queries to human agents—reflects the need for careful oversight of confidential matters. Results must be explainable and perceived as fair in all cases.

Effective knowledge management is another requirement. Revenue authorities have extensive laws, regulations, case records, and operational manuals. However, scattered archives and incomplete digitization can hamper efforts to train AI systems effectively. A human must determine which documents are accurate, relevant, and suitable for inclusion in the training material.

As GenAI becomes integrated into various aspects of revenue administration, employees will need to be trained to interpret, correct, and complement its outputs. Policymakers must ensure that errors are reported and addressed promptly.

By providing human-like capabilities to support taxpayers and tax authorities, GenAI can act as both taxman and taxpayer assistant, automating routine tasks, clarifying complex issues, and fostering a more transparent and collaborative relationship.

This technology can lower administrative hurdles, demystify tax obligations, and invite broader participation in policy debates. However, shaping it properly requires strong leadership, ethical policy frameworks, and vigilant oversight of data quality, privacy, and accuracy.

Thomas Cantens is head of research and policy unit at the World Customs Organization; Herve Tourpe is Chief of Digital Advisory Unit, International Monetary Fund (IMF).

IPS UN Bureau

Follow @IPSNewsUNBureau
  

 
CIVICUS speaks with Olivia Sohr about the challenges of disinformation and the consequences of the closure of Meta’s fact-checking programme in the USA. Olivia is the Director of Impact and New Initiatives at Chequeado, an Argentine civil society organisation working since 2010 to improve the quality of public debate through fact-checking, combating disinformation, promoting access to information and open data.

Olivia Sohr

What led to Meta’s decision to end its fact-checking programme?

While the exact details of the process that led to this decision are unknown, in his announcement Zuckerberg alluded to a ‘cultural shift’ that he said was cemented in the recent US election. He also expressed concern that the fact-checking system had contributed to what he saw as an environment of ‘excessive censorship’. As an alternative, Zuckerberg is proposing a community rating system to identify fake content.

This decision is a setback for information integrity around the world. Worryingly, Meta justifies its position by equating fact-checking journalism with censorship. Fact-checking is not censorship; it’s a tool that provides data and context to enable people to make informed decisions in an environment where disinformation is rife. Decisions like this increase opacity and hamper the work of those focused on combatting disinformation.

The role of fact-checkers in Meta is to investigate and label content that is found to be false or misleading. However, decisions about the visibility or reach of such content will be made solely by the platform, which has assured that it will only reduce exposure and add context, not remove or censor content.

How the community grading system will work has not yet been specified, but the prospects are not promising. Experience from other platforms suggests that these models tend to increase disinformation and the spread of other harmful content.

What are the challenges of fact-checking journalism?

Fact-checking is extremely challenging. While those pushing disinformation can quickly create and spread completely false content designed to manipulate emotions, fact-checkers must follow a rigorous and transparent process that is time-consuming. They must constantly adapt to new and increasingly sophisticated disinformation strategies and techniques, which are proliferating through the use of artificial intelligence.

Meta’s decision to end its US verification programme makes our task even more difficult. One of the key benefits of this programme is that it has allowed us to reach out directly to those who spread disinformation, alerting them with verified information and stopping the spread at the source. Losing this tool would be a major setback in the fight against disinformation.

What are the potential consequences of this change?

Meta’s policy change could significantly weaken the information ecosystem, making it easier for disinformation and other harmful content to reach a wider audience. For Chequeado, this means we will have to step up our efforts to counter disinformation, within the platform and in other spaces.

In this scenario, verification journalism is essential, but it will be necessary to complement this work with media literacy initiatives, the promotion of critical thinking, the implementation of technological tools to streamline the work and research to identify patterns of disinformation and the vulnerability of different groups to fake news.

GET IN TOUCH
Website
Instagram
Twitter

SEE ALSO
BRAZIL: ‘The focus should be on holding social media companies accountable, not punishing individual users’ CIVICUS Lens | Interview with Iná Jost 01.Oct.2024

‘It’s easier and cheaper than ever to spread disinformation on a massive scale’ CIVICUS Lens | Interview with Imran Ahmed 21.Sep.2024

UK: ‘Social media platforms have become breeding grounds for far-right ideologies’ CIVICUS Lens | Interview with Kulvinder Nagre 19.Aug.2024

Follow @IPSNewsUNBureau
  

Women in Kilema village harvest orange sweet potatoes. Credit: Kizito Makoye/IPS

By Kizito Makoye
KILIMANJARO, Tanzania , Dec 4 2024 (IPS)

In the sun-scorched soils of Moshi, where every drop of rain counts, two female farmers have defied the odds through technology. Mwajuma Rashid Njau and Mumii Rajab, once locked in a daily struggle to survive, have found a mobile phone their best ally.

For years, farming was a way of life they struggled to master. Their fields, a patchwork of red earth and wilting crops, symbolized hardship rather than prosperity. Pests came with the seasons, the soil quality deteriorated, and their harvests barely provided enough to feed their families. But now, a simple app—Kiazi Bora—has changed everything.

On a sweltering afternoon, Njau was out in the field, staring helplessly at the rows of wilting sweet potatoes ravaged by pests, when he realized things could be different. She had no idea how to stop it—until she opened the Kiazi Bora app on her phone.

“This app has changed everything,” Njau, 38, says with a tired but hopeful smile. “I didn’t know where to start, but now I can check my phone, and it tells me exactly what to do.”

The Kiazi Bora app, designed specifically for small-scale farmers like Njau and Rashid, focuses on helping them grow nutritious orange-fleshed sweet potatoes (OFSP) to feed their families and earn income. The app offers simple instructions on planting and pest control to farmers with little education.

The app, Kiazi Bora (“quality potatoes” in Kiswahili), wasn’t just another farming tool—it was powered by cutting-edge AI voice technology. And for the first time, it spoke their language.

Creating Kiazi Bora wasn’t easy. Kiswahili, a language spoken by over 200 million people, presented unique challenges for AI developers. The problem? There simply wasn’t enough high-quality voice data to train the technology.

“One of the biggest challenges has been the availability of diverse, high-quality data,” said EM Lewis-Jong, Director of Mozilla Common Voice, a global project dedicated to making AI accessible to speakers of underrepresented languages.

“Kiswahili is a diverse language with many regional variants, and our tools are primarily designed for English, which complicates things further.”

To solve this issue, SEE Africa, the nonprofit behind Kiazi Bora, turned to Mozilla’s Common Voice platform. Unlike other AI data collection methods, which often rely on scraping the web or underpaid gig workers, Common Voice harnesses the power of community. “We use a crowd-sourced model where people voluntarily contribute their voice data,” explained Lewis-Jong. “This ensures that the data reflects the true diversity of the language, including different accents and dialects.”

This community-driven approach has already seen tremendous success. In Tanzania, the Kiazi Bora app is now used by over 300 women, empowering them with knowledge on how to grow and market their crops. “These women are learning in Kiswahili, their first language, which makes a huge difference,” noted Gina Moape, Community Manager for Common Voice. “We’ve seen firsthand how access to information in their own language improves both their nutrition and their ability to participate in economic activities.”

But Kiazi Bora is just one example of how voice-enabled technology can make a real impact.

For Mozilla, these projects reflect a broader vision: democratizing AI so that it serves everyone, not just speakers of dominant languages. “If data creation is left to for-profit companies, many of the world’s languages will be left behind,” said Lewis-Jong. “We want a world where people can create the data they need, capturing their language as they experience it.”

That’s why Mozilla’s Common Voice is not just a tool but a movement. Its open-source platform allows communities to collect and contribute voice data that anyone can use, fostering local innovation across Africa. “We’re particularly excited about the potential for African languages,” Lewis-Jong added. “Our long-term vision is to integrate more African languages into global voice recognition technologies, and Common Voice is a critical part of making that happen.”

For Rashid, 42, who had once lived in uncertainty, the app was a useful tool. “Before, I felt powerless,” she recalls. “When pests attacked, I would just watch as my crops withered. Now, I can fight back. I know what to do.”

Both women have honed their skills and improved crop yields. The app taught them how to manage soil health, optimize planting schedules, and handle pest outbreaks.

Their orange-fleshed sweet potatoes stand out in contrast to the dusty earth, a sign of resilience and renewal.

The duo, who were entangled in a cycle of poverty, now speak with pride about their success.

“We’ve learned to control our future,” Njau says.

Through Kiazi Bora, Njau and Rajabu have unlocked opportunities to improve their livelihoods and break free from poverty.

Njau, who had to drop out of school when her family moved to a remote village, calls the app her “teacher.” She explains, “I never completed school, but this app has taught me everything I need to know about farming. It’s like a teacher that’s always there when I need it.”

The voice-enabled Kiswahili features make it user-friendly. “The app speaks to me in a language I clearly understand,” Njau says.

Through the app, Njau and Rajabu learned how to process potatoes into flour and pastries, which fetch a higher market price.

Rajabu explains, “I didn’t know you could make flour from sweet potatoes or that you could sell it for more money. Now, I have customers who buy the flour because it lasts longer than fresh potatoes.” This new skill has allowed them to diversify their income.

In just a year, their income increased from zero to USD 127 per month. The extra income has enabled them to take care of their families, reinvest in their farms, and secure a better future. “With the money I’ve made, I’ve been able to send my children to school and even save some for emergencies,” says Njau.

The potatoes, which are rich in vitamins, have helped them fight malnutrition in their communities. While neither Njau nor Rajabu had children with malnutrition, they both knew families who struggled with it. Thanks to the app, they now understand the importance of incorporating OFSP into their daily meals to ensure their children stay healthy.

Rajabu was quick to share the app with her relatives. “I told my sister about it, and now she’s also growing OFSP. Her children are healthier, and she’s even making money from selling sweet potato flour,” she says proudly.

For both women, the app has empowered them as farmers, businesswomen and community leaders. “I feel confident now,” Rajabu says. “This app has changed my life, and I know it can help other women like me.”

Both Njau and Rajabu see immense potential for Kiazi Bora to help other rural women. They advocate for expanding the app beyond OFSP farming to include other crops like vegetables and edible roots, as this could further diversify their income streams and enhance food security in their communities.

“Women in rural areas need this technology,” Rajabu emphasizes. “We need to make sure that we can feed our families and earn better incomes.”

IPS UN Bureau Report

Follow @IPSNewsUNBureau