Artificial intelligence Archives | Page 26 of 676

European Commission advances AI transparency code under EU AI Act

The European Commission’s AI Office has convened a new round of working group meetings and workshops on the forthcoming Code of Practice on Marking and Labelling of AI-Generated Content.

The discussions brought together providers of generative AI systems and models, technology companies, industry representatives, civil society organisations and academic experts. Feedback from the meetings will inform the third and final draft of the code, expected in early June.

The code is intended to support transparency obligations under the AI Act, including requirements linked to marking, labelling, disclosure and detectability of AI-generated content. It covers issues such as synthetic media, deepfakes and certain AI-generated text.

Working Group 1 focused on marking and detection obligations for providers, including a revised multi-layered approach, technical feasibility, benchmarking, compliance frameworks and possible third-party assessments. Industry participants raised concerns over compliance burdens, innovation and feasibility, while civil society and academic experts called for stronger safeguards in the public interest.

Working Group 2 examined disclosure obligations for deployers of generative AI systems, particularly deepfakes and certain AI-generated text. Discussions covered origin disclosure, user-facing labels, proportionality, governance measures, editorial control and the possible development of a uniform EU label.

Additional workshops explored how machine-readable marks, provenance data, visible labels, watermarking systems and an EU-wide icon could work together across the AI value chain. Participants also discussed coordination with other EU rules, including the Digital Services Act, while stressing the need to balance transparency, legal clarity, accessibility and innovation.

Why does it matter?

The code of practice will help determine how AI-generated content is marked, labelled and disclosed across the EU. Its development highlights the practical difficulty of turning transparency obligations into workable rules, particularly when regulators, companies and civil society disagree over technical feasibility, compliance costs, user experience and safeguards against deceptive synthetic media.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Grokipedia articles show selective political divergence from Wikipedia, research finds

A new study published in the Proceedings of the National Academy of Sciences examined structural and political differences between Wikipedia and Grokipedia, the AI-generated encyclopedia developed by xAI.

Researchers analysed 17,790 matched article pairs drawn from the 20,000 most-edited English-language Wikipedia entries. They found that Grokipedia articles are typically longer, more syntactically complex, and contain fewer references and hyperlinks per 1,000 words than their Wikipedia counterparts.

The study also identified a bimodal pattern across similarity measures, indicating that some Grokipedia entries closely resemble Wikipedia entries, while others diverge substantially in content and structure. Researchers said the findings suggest Grokipedia is not a fully independent alternative to Wikipedia, but often appears as an AI-mediated reconfiguration of Wikipedia content.

The analysis examined ideological differences by evaluating the political orientation of cited news media sources. Researchers found that divergence was concentrated primarily in politically and culturally sensitive topics, including religion, history, politics and literature.

Within those areas, Grokipedia articles showed a relative shift toward more right-leaning cited sources than Wikipedia. However, the study also noted that sources cited on both platforms remained predominantly left-leaning.

Researchers argued that Wikipedia’s human editorial processes make disputes, revisions and bias visible and contestable, while AI-generated systems may embed bias within more opaque automated workflows that are harder to scrutinise publicly.

The paper also raised broader concerns about the governance of AI-generated knowledge systems. Researchers warned that AI-generated encyclopedic content could shape future training datasets and automated information ecosystems, potentially reproducing or amplifying bias without sufficient transparency, accountability or human oversight.

Why does it matter?

The findings add to growing debates over AI-generated knowledge systems, political bias, citation quality and transparency. As generative AI increasingly produces reference and educational material, the key question is not only whether outputs are accurate, but whether their sources, editorial assumptions and revisions can be scrutinised. Grokipedia’s differences from Wikipedia show how automated knowledge systems may reshape information governance while making some forms of bias less visible.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

European Central Bank examines systemic risks linked to AI-driven finance

The European Central Bank (ECB) has published research examining how AI systems could affect financial stability as AI adoption expands across financial markets.

According to Eurosystem research, different AI architectures may produce significantly different market behaviours under similar economic conditions.

ECB simulations compared reinforcement learning systems with large language model-based agents operating in simulated financial environments. Researchers found that some reinforcement learning systems displayed coordinated responses resembling bank run dynamics in certain scenarios.

The report linked part of this behaviour to risk-avoidance patterns associated with prior negative outcomes.

Large language model-based systems showed lower coordination but more variable and unpredictable responses during periods of uncertainty. Despite receiving identical instructions, LLM-based agents frequently developed different assumptions about market behaviour, particularly during periods of moderate economic uncertainty.

ECB researchers noted that such inconsistency could create its own form of instability as AI-generated expectations diverge across financial markets.

The ECB suggested that wider AI adoption in finance may require updated risk management practices, investor awareness, and regulatory safeguards.

The research also highlighted the potential importance of existing market stabilisation measures, including circuit breakers and investor protection mechanisms.

Why does it matter?

AI is rapidly becoming embedded in trading, investment management, and financial decision-making across global markets, meaning flaws in AI behaviour could amplify systemic risks at unprecedented speed and scale.

The research signals that financial stability may increasingly depend not only on economic fundamentals and regulation, but also on the underlying architecture, coordination patterns, and predictability of the AI systems shaping market activity.

Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our chatbot!

South Korea launches tripartite committee on AI and labour

South Korea’s Economic, Social and Labor Council (ESLC), a presidential advisory body, has launched a tripartite committee to examine the impact of AI on labour and workplaces. The committee brings together labour, business, government, and public interest representatives for a year-long dialogue on AI-related workplace changes.

The committee held its first meeting in Seoul and will examine how AI adoption may affect employment patterns and industrial workplaces. The 17-member body is chaired by former presidential jobs secretary Hwang Deok-soon and includes labour, business, government and public interest representatives.

According to the ESLC, discussions will focus on AI adoption, changes to jobs and work tasks, worker data collection, and support measures linked to workforce transitions. The initiative is expected to include expert consultations and workplace assessments examining practical uses of AI technologies.

The launch comes amid broader public debate about automation, humanoid robotics, and potential labour-market disruption linked to AI technologies. ESLC Chair Kim Ji-hyung said the discussions aim to balance technological development, industrial competitiveness, and labour market stability in South Korea.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

World Bank highlights ‘Small AI’ for farmers and rural communities

According to Hindustan Times, World Bank President Ajay Banga highlighted the potential of ‘Small AI’ systems to support farmers and rural communities through locally deployed and lower-cost technologies.

Examples discussed included farmers in India using mobile phones to share images of diseased crops and receive agricultural advice remotely. Banga also referred to healthcare workers in Indonesia using basic internet connections to access local diagnostic support systems in remote areas.

At the summit, entrepreneur Saurav Mukherjee said AI adoption was expanding into sectors including agriculture and food production. Mukherjee said AI tools may support agricultural decision-making through analysis of seed quality and environmental conditions such as soil, weather, and water availability.

He also noted that wider internet connectivity and 5G access could support wider AI adoption in underserved regions. However, he cautioned that shortages of skilled workers could limit implementation capacity in some regions.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Ofcom says major platforms will introduce new child safety measures in the UK

Ofcom said Snap, Meta, and Roblox plan to introduce additional child safety measures in the UK following regulatory pressure to strengthen protections against online grooming.

The commitments were outlined in an Ofcom report reviewing how major platforms responded to requests for stronger child safety protections. The measures include changes to contact settings, age assurance systems, AI-assisted detection tools, and controls for direct messaging.

According to Ofcom, Snap committed to implementing all recommended grooming prevention measures under the regulator’s Illegal Harms Codes. The changes will prevent adult strangers from contacting children by default, prevent children from being encouraged to expand their friendship groups to people they do not know, and introduce stronger age checks for UK users over the summer.

Roblox said it would expand existing protections, including parental controls for direct chat functions involving younger users. Meta said it plans to introduce additional privacy settings for teen accounts and AI-assisted detection tools for potentially inappropriate interactions.

Ofcom said it remains concerned about measures related to recommendation systems and harmful content exposure on some platforms. The regulator said evidence shows children’s exposure to harmful content has changed little since online safety duties came into force in July 2025.

The regulator also said platforms have not yet demonstrated effective enforcement of minimum age requirements. Its research found that 84% of children aged 8 to 12 were still using one of the top five online services, despite minimum age rules.

Ofcom has written to the UK Secretary of State advising that online safety legislation would need a clearer basis if Parliament wants the regulator to force platforms to enforce minimum age policies effectively.

Ofcom’s action plan includes monitoring implementation, reviewing recommender system risks, considering enforcement action, and continuing research into children’s online experiences.

Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!

Canada and Spain sign AI cooperation agreement

Canada and Spain have signed a memorandum of understanding aimed at strengthening cooperation on AI development, adoption, and digital innovation.

The agreement was signed in Toronto by Canada’s Minister of AI and Digital Innovation Evan Solomon and Spanish Deputy Prime Minister Carlos Cuerpo during a visit by King Felipe VI of Spain.

The memorandum establishes a framework for cooperation on AI technologies, computing capacity, and collaboration between governments, industry, and other stakeholders. According to both governments, the partnership aims to support innovation, investment, and economic growth linked to AI technologies.

Discussions during the visit focused on innovation ecosystems, AI adoption by businesses, and opportunities for small and medium-sized enterprises.

Participants also discussed the role of public policy and private-sector cooperation in supporting trustworthy AI and technology deployment.

Solomon said: ‘Today’s engagement with Spain reflects the importance of trusted international partnerships in shaping how artificial intelligence develops in a safe and responsible way. By working together, we can better align innovation with shared values, strengthen research ties and ensure AI delivers long-term economic opportunity and tangible benefits for both countries.’

Cuerpo said: ‘This agreement opens a new chapter in the relationship between Spain and Canada. Artificial intelligence is one of the forces that will shape competitiveness and economic sovereignty in the decades ahead, and two democracies such as ours have a responsibility to lead together, building a model of technological cooperation grounded in trust, security and shared benefit.’

The agreement reflects broader international efforts to strengthen cooperation on AI governance, research, and economic development.

Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!

Singapore launches new AI, cybersecurity and quantum-readiness programmes

Singapore has announced new initiatives aimed at supporting enterprise AI adoption, strengthening cybersecurity, and preparing digital infrastructure for future quantum-related risks.

The measures were announced at ATxEnterprise 2026 by Senior Minister of State for Digital Development and Information Tan Kiat How. They include new partnerships under the Digital Enterprise Blueprint, an AI adoption playbook for enterprises, SME awards recognising AI impact, and a pilot on quantum-safe technologies.

According to IMDA’s Singapore Digital Economy Report 2025, AI adoption among SMEs increased significantly during 2024.

IMDA and the Singapore Business Federation will introduce SME AI Impact Awards recognising enterprises using AI technologies in business operations. Up to 30 winners will be recognised across categories for proprietary AI tools and adoption of ready-to-use AI solutions.

The Digital Enterprise Blueprint is being expanded through partnerships involving AI training, digital skills development, and cybersecurity support for SMEs. One programme led by Grab will provide AI-related training and courses for SMEs in sectors including retail, e-commerce, and food services.

RSM Stone Forest IT will also launch cybersecurity initiatives involving phishing simulations, awareness webinars, and tabletop exercises for SMEs. With the two partnerships, IMDA aims to reach 12,000 more SMEs, contributing to its target of supporting 50,000 SMEs by 2029.

IMDA, SkillsFuture Singapore, and Workforce Singapore have also developed an AI for Enterprise Impact Playbook to help digitally progressive enterprises assess readiness, identify support, and plan next steps for AI adoption.

Singapore additionally announced a pilot initiative focused on quantum-safe technologies for telecommunications infrastructure. IMDA signed a Memorandum of Intent with Singtel, Ericsson, and NCS Singapore to test and validate quantum-safe migration approaches.

Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!

Ofcom report highlights growing AI use among UK children online

The UK’s Ofcom has released new research indicating that children in the UK are using digital devices and online services at increasingly younger ages.

According to Ofcom’s Children’s Online Experiences report, screen use begins early in childhood, and smartphone ownership increases significantly during secondary school years. The report found that teenagers aged 15 to 17 spend a substantial amount of time online each week.

The report also noted declining use of traditional media formats such as live television, radio, and print among younger audiences. Live television, radio, and print media were described as increasingly absent from children’s routines, with social media, messaging platforms, and gaming dominating digital engagement.

Ofcom also warned that exposure to harmful content remains a significant issue despite the introduction of new online safety rules. Ofcom said many children reported exposure to harmful online content, including material surfaced through recommendation systems and personalised feeds.

The report also highlighted growing use of AI tools among children and teenagers. More than half of UK children aged 8 to 17 said they use AI tools, with some teenagers increasingly relying on AI systems for learning, creativity, communication, and companionship. Researchers said some children found it difficult to distinguish between AI-generated and human-created content.

The report suggested that passive content consumption plays an increasingly significant role in children’s online activity. Most younger users primarily scroll, watch, follow, or like content instead of actively creating or sharing material themselves.

Gaming remained one of the most important online social environments for children, with many users interacting regularly with people they had only met online through multiplayer gaming communities and communication platforms.

Why does it matter?

Ofcom’s findings highlight growing concerns surrounding children’s digital well-being, algorithmic exposure, AI literacy, and online safety regulation. Policymakers and regulators increasingly face pressure to address how recommendation systems, generative AI, and social platforms shape behaviour, attention, and trust among younger audiences.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

International Labour Organization warns AI could reshape labour markets across the Arab region

The International Labour Organization (ILO) and the United Nations Economic and Social Commission for Western Asia (ESCWA) have examined how AI may reshape labour markets and employment patterns across the Arab region.

The organisations released a report exploring how AI adoption may transform jobs, productivity, and workforce dynamics by 2035. According to the report, outcomes will depend on policy choices related to skills development, labour protections, and social support systems.

The report outlines three possible scenarios ranging from inclusive AI-driven growth to increased inequality linked to insufficient labour protections and workforce adaptation measures.

One projected strong AI-driven economic growth, combined with large-scale investment in workforce transition and retraining programmes.

Another warned that rapid technological adoption without sufficient social safeguards could deepen inequality and displace large numbers of lower- and middle-skilled workers.

A third scenario envisaged a more gradual AI integration, supported by coordinated policy reforms and inclusive labour-market strategies.

The report identifies sectors such as healthcare, education, logistics, tourism, and digital services as areas where AI-related employment opportunities may emerge. At the same time, the organisations noted that automation could reduce demand for some routine and clerical occupations.

ILO Regional Director for Arab States Ruba Jaradat said AI technologies are already affecting workplaces across public administration and service sectors in the region. She added that nearly one-quarter of occupations may experience either displacement or technological augmentation linked to generative AI systems.

The analysis also highlighted widening skills mismatches between education systems and labour market demands, with some countries facing gaps ranging from 40% to 70%. The report also highlights the importance of investment in lifelong learning, labour market institutions, social protection, and AI governance frameworks.

The discussions took place during a preparatory session linked to the Arab Forum for Sustainable Development, where policymakers, labour organisations, and international experts examined how AI may affect youth employment, women workers, and lower-skilled populations across the region.

Why does it matter?

ILO highlights how developing and emerging economies may experience AI transitions differently depending on infrastructure, education systems, governance capacity, and investment levels. Policymakers across the Arab region are now under increasing pressure to modernise labour systems while ensuring that AI adoption supports inclusive growth instead of deepening social inequality.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!