Artificial intelligence Archives | Page 10 of 699

China adopts national standards for AI agents

China’s State Administration for Market Regulation has released seven national standards for AI agent interconnection, establishing a common framework for how autonomous AI systems identify themselves, communicate and operate across platforms and industries.

The standards define AI agents as intelligent systems capable of autonomous perception, memory, decision-making, interaction and execution. The administration described them as a key application of next-generation AI and an important mechanism for deploying AI capabilities across industries.

By standardising architecture and interaction rules, the framework aims to help companies reuse common components, reduce customised development and shorten the time needed to bring AI-powered products to market.

The standards also introduce unified identity authentication and end-to-end traceability mechanisms, addressing what the regulator described as a significant gap in existing governance of AI agent systems.

Why does it matter?

AI agents will increasingly need to interact with other systems, services and organisations rather than operate in isolation. Common technical standards can improve interoperability, reduce development costs and make it easier for businesses to deploy AI applications across different sectors and platforms.

The standards also illustrate China’s strategy of shaping emerging AI technologies through nationally coordinated technical frameworks. By establishing common rules for identity, interaction and traceability at an early stage of the technology’s development, China is positioning itself to influence how agentic AI ecosystems evolve domestically and potentially internationally.

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

Anthropic details Claude Fable 5 cyber safeguards

Anthropic has released new details about the cybersecurity safeguards protecting Claude Fable 5 and proposed a draft framework for assessing the severity of AI cyber jailbreaks.

The company said Fable 5 has now been redeployed globally and is available to all users. The update focuses on two areas: safety classifiers designed to detect and block dangerous cybersecurity activity, and a proposed framework for grading the severity of cyber-related AI jailbreaks.

Anthropic said Fable 5 uses safety classifiers to distinguish between four categories of cybersecurity activity: prohibited use, high-risk dual use, low-risk dual use and benign use. Prohibited activities include destructive cyberattacks, cyber-physical sabotage, defence evasion, command-and-control infrastructure, data exfiltration, malware development and attacks targeting internet backbone systems.

High-risk dual-use activities include penetration testing, red teaming, privilege escalation, exploit development, virtual machine escapes and advanced vulnerability discovery. Anthropic said these activities will remain restricted for Fable 5 until stronger access controls for trusted users are in place.

The proposed Cyber Jailbreak Severity (CJS) framework introduces a scale ranging from CJS-0 to CJS-4, assessing jailbreaks according to capability gain, breadth of capability gain, ease of weaponisation and discoverability.

Anthropic said the framework is intended to give AI developers, governments and security researchers a common vocabulary for describing cyber jailbreak risks. The company is seeking public feedback on the draft and has launched a HackerOne programme inviting researchers to submit potential cyber jailbreaks affecting Fable 5.

Why does it matter?

As AI models become more capable in cybersecurity, assessing jailbreaks consistently is becoming increasingly important. A common severity framework could help AI developers, researchers, and governments compare vulnerabilities, prioritise responses and communicate risks using shared criteria rather than ad hoc descriptions.

The proposal also reflects a broader shift toward treating frontier AI safety as a collaborative effort. By publishing its methodology and inviting external researchers to test Fable 5 through a coordinated disclosure programme, Anthropic is encouraging greater transparency and independent scrutiny of AI security safeguards.

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

India backs ‘Human-led AI’ for digital government

India has reaffirmed its commitment to placing people at the centre of AI during the 29th National Conference on e-Governance, where Union Minister Dr Jitendra Singh said ‘Human-led Artificial Intelligence’ will guide the country’s journey towards Viksit Bharat 2047.

Speaking at the conference’s closing session in Jaipur, Singh said AI is no longer optional for governments, but argued that its success depends on responsible deployment that strengthens transparency, accountability and public service rather than replacing human judgement.

Held under the theme ‘Viksit Bharat 2047: AI-enabled, Data-driven and Secure Digital Governance’, the conference brought together policymakers, technology experts, researchers, industry representatives and public administrators to discuss the future of digital government.

The conference concluded with the adoption of the Jaipur Declaration, which sets out a strategic roadmap for AI-enabled, secure and citizen-centric governance. It also recognised 17 digital governance initiatives through the National e-Governance Awards 2026, highlighting innovation across ministries, states, local governments and research institutions.

Throughout the event, speakers presented AI as a tool to strengthen public administration while preserving democratic accountability. Singh stressed that AI should enhance institutional capacity rather than replace human responsibility.

He highlighted several flagship initiatives, including the Centralized Public Grievance Redress and Monitoring System, the multilingual AI chatbot SAMADHAN DIDI developed with BHASHINI, the National e-Governance Service Delivery Assessment, Mission Karmayogi and the IndiaAI Mission, all intended to improve efficiency, accessibility and citizen engagement.

Singh also said India’s long-term digital strategy extends beyond technology deployment. Capacity building for civil servants, administrative reform, secure Digital Public Infrastructure and responsible AI governance are all seen as essential to achieving the country’s 2047 development ambitions. According to Singh, AI should accelerate public service delivery while remaining grounded in ethics, constitutional values and human oversight.

According to Dr Singh, technology should accelerate governance while remaining firmly guided by ethics, constitutional values and human oversight.

Why does it matter?

India is positioning AI as a tool to strengthen public administration rather than replace human decision-making. By emphasising human oversight, ethics and citizen-centred services, the government is seeking to balance technological innovation with democratic accountability as AI becomes more deeply integrated into public institutions.

The Jaipur Declaration also signals that AI is becoming a long-term pillar of India’s digital governance strategy. Combined with investments in Digital Public Infrastructure, civil service capacity development and multilingual AI services, the approach could shape how other countries integrate AI into public administration while maintaining public trust.

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

Bank of England explores AI ‘kill switches’ for markets

The Bank of England is exploring whether emergency ‘kill switches’ could halt trading activity if autonomous AI systems begin behaving unpredictably, reflecting growing concern that existing market safeguards may not be suited to AI-driven trading.

Deputy Governor Sarah Breeden said regulatory frameworks must evolve as AI systems become capable of autonomously chaining actions and executing trading strategies. Speaking at the European Central Bank’s Sintra Forum, she warned that relying solely on human oversight may no longer be sufficient as financial markets become increasingly automated.

Regulators are particularly concerned about systemic risks, including AI models amplifying market volatility, exhibiting coordinated behaviour or pursuing objectives that diverge from their intended design. They also warned that AI could strengthen cyber defences while simultaneously making it easier to discover and exploit vulnerabilities at scale.

Breeden said the Bank is also exploring resilience measures such as simulation testing, stronger recovery mechanisms and potential cross-institution support during market disruptions. She added that international coordination will be essential as increasingly autonomous AI systems become embedded in global financial markets.

Why does it matter?

The Bank of England’s proposals reflect a growing recognition that autonomous AI systems could introduce systemic risks that existing market safeguards were not designed to address. Traditional mechanisms such as circuit breakers assume markets are ultimately driven by human decisions, whereas AI agents may react to changing conditions at machine speed and in highly coordinated ways.

The discussion also illustrates how financial regulators are shifting from studying AI risks to preparing practical resilience measures. Tools such as simulation testing, emergency trading controls and international coordination could become increasingly important as AI takes on a larger role in trading, payments and other core financial market functions.

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

UN welcomes first global scientific assessment of AI

The United Nations has welcomed the publication of the first independent global scientific assessment of AI, describing it as a landmark effort to evaluate the technology’s opportunities, risks and societal impacts across multiple domains.

Produced by a 40-member international expert panel, the report examines AI’s impact on healthcare, education, the economy, security, human rights and governance. It highlights rapid advances in areas such as medical research and productivity while warning of growing concerns over safety, inequality and regulatory gaps.

UN Secretary-General António Guterres said the assessment leaves little doubt about AI’s transformative impact, urging governments to strengthen governance frameworks without delay. He warned that the benefits of AI will not be shared equitably without coordinated international actions.

The experts also cautioned that AI capabilities are advancing faster than scientific understanding and regulatory capacity. While AI could accelerate progress across multiple sectors, they warned of growing risks including misinformation, cybersecurity threats, labour market disruption and unequal access to the technology.

Why does it matter?

The assessment establishes a common scientific reference point for discussions on AI governance at a time when countries are pursuing increasingly different regulatory approaches. By providing an independent, multidisciplinary evaluation of AI’s opportunities and risks, it could help inform future international cooperation and policymaking.

The report also reinforces the growing gap between the pace of AI development and the capacity of governments and institutions to respond. As AI becomes more deeply embedded across economies and public services, coordinated governance, investment in oversight, and international collaboration will be increasingly important to ensure the technology’s benefits are broadly shared while its risks are effectively managed.

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

AI for Good Global Commission launches to expand trusted AI access

Rwanda’s President Paul Kagame, Salesforce Chair and CEO Marc Benioff and International Telecommunication Union Secretary-General Doreen Bogdan-Martin have announced the launch of the AI for Good Global Commission.

The Commission brings together more than 40 founding members, including heads of state and government, technology executives and heads of UN agencies.

It is co-chaired by Kagame and Benioff, with Bogdan-Martin serving as vice-chair. ITU said the Commission will work to identify practical pathways to strengthen trust, expand access and unlock AI’s potential to address real-world challenges.

The initiative will focus on technical, socioeconomic and policy questions around AI, with an emphasis on responsible innovation, human capability and broad-based economic and social benefits.

Access is a central part of the Commission’s mandate. ITU said 2.2 billion people remain offline, limiting their ability to benefit from AI developments.

The Commission builds on ITU/UNESCO Broadband Commission for Sustainable Development, which has focused on connectivity, digital inclusion and economic development.

Its inaugural meeting will take place during the AI for Good Global Summit 2026 from 7 to 10 July. The Summit is part of Digital Week, which also includes the first UN-mandated Global Dialogue on AI Governance and the WSIS Forum 2026.

Why does it matter?

The AI for Good Global Commission places digital inclusion at the centre of global AI governance debates. Its launch highlights a key challenge: many countries and communities cannot benefit from AI if they lack connectivity, infrastructure, skills and institutional capacity. The Commission’s relevance will depend on whether it can move beyond high-level commitments and help turn access, trust and responsible innovation into practical support for developing countries.

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

Intesa Sanpaolo moves core IT systems to Google Cloud in Italy

Intesa Sanpaolo has completed the migration of core IT systems to Google Cloud’s Italian regions, hosted in TIM data centres in Milan and Turin.

The Italian banking group said the project is one of the largest cloud migrations carried out by a European financial institution. It forms part of a broader digital transformation effort with Google Cloud and TIM.

The migration involved more than 800 business applications and the retirement of a similar number of legacy systems previously hosted on the bank’s own infrastructure.

According to the partners, the transition was completed while maintaining operational continuity for the bank’s services.

Google Cloud provided cloud infrastructure and data capabilities, while TIM hosted the Italian cloud regions and supported data centre services, connectivity and project governance.

Intesa Sanpaolo said the migration strengthens its cloud-first strategy and provides the technological foundation for Isytech, its cloud-native digital platform for customers and employees across the group.

The bank also linked the project to workforce transformation. More than 3,000 employees participated in cloud training programmes, earning more than 170 Google Cloud certifications.

The project is intended to support advanced analytics, AI adoption and future digital banking services while keeping workloads in Google Cloud’s Italian regions.

Why does it matter?

The migration shows how major European banks are modernising legacy IT infrastructure to support AI, analytics and digital financial services. Moving core systems to cloud environments can improve scalability and accelerate new services, but it also raises important questions about resilience, cybersecurity, data governance and regulatory compliance. The use of Italian cloud regions hosted in TIM data centres reflects the growing importance of data residency and trusted infrastructure in regulated financial services.

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

India and Japan expand strategic AI partnership

India and Japan have agreed to deepen cooperation on AI, linking AI governance, cybersecurity, infrastructure, research and talent development.

In a joint statement, the two countries described AI as a transformative technology with long-term implications for innovation, economic security, governance and the international order.

Both sides are committed to building a safe, secure, trustworthy, inclusive and human-centric AI ecosystem. They also agreed to strengthen cooperation with partners in the Indo-Pacific and the Global South.

The statement identifies international AI governance, safety and cybersecurity as priority areas. India and Japan said they would coordinate in forums including the G20, OECD, Global Partnership on AI and the UN, while supporting responsible innovation and risk-based governance.

The two countries also agreed to cooperate on AI-enabled cybersecurity and the security of AI systems, with particular attention to critical infrastructure. They highlighted the need for safeguards to ensure AI supports children’s learning and growth rather than causing harm.

AI infrastructure is another focus. India and Japan will strengthen cooperation on data centres, GPU and other compute resources, semiconductors and trustworthy supply chains across the AI technology stack.

The statement also supports collaboration on multilingual, open-source and domain-specific AI models, including models for native languages and public-interest applications. Several memoranda were signed, including partnerships involving IIT Bombay, BharatGen, Japan’s National Institute of Informatics, Sarvam, Preferred Networks, IndiaAI and Japan’s Ministry of Economy, Trade and Industry.

Both sides also committed to researcher exchanges, industry-academia collaboration and talent mobility. Japan reaffirmed its goal of welcoming 500 highly skilled AI professionals from India by 2030.

Why does it matter?

The joint statement shows how AI cooperation is becoming part of broader economic and security strategies in the Indo-Pacific. India and Japan are not only discussing AI governance, but also the infrastructure and supply chains needed to build and deploy AI systems, including compute, semiconductors, data centres and talent. The focus on multilingual and open-source models also matters for countries seeking AI systems that reflect local languages, public-interest needs and Global South priorities.

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

Ireland introduces AI Bill to implement EU AI Act

Ireland’s government has introduced the Regulation of AI Bill 2026, with Digital Transformation Minister Niamh Smyth describing the legislation as essential to protecting citizens while supporting innovation during its Second Stage debate in the Dáil.

The Bill is intended to give full effect to the EU AI Act in Ireland by establishing the national institutions needed to supervise and enforce the regulation ahead of the EU implementation deadline of 2 August 2026.

A central element of the Bill is the establishment of the AI Office of Ireland as an independent statutory body. The office will act as Ireland’s national point of contact with the European Commission and other member states, oversee enforcement of the AI Act, promote AI literacy and innovation, and operate a regulatory sandbox for start-ups and small and medium-sized enterprises.

Smyth acknowledged both the opportunities and risks presented by AI, highlighting advances in areas such as healthcare and scientific research while warning that, without appropriate safeguards, the technology could reinforce discrimination, manipulate behaviour and exploit vulnerabilities. She emphasised that the Bill is an implementing measure and does not introduce obligations beyond those already established by the EU AI Act.

Smyth also said the legislation would strengthen Ireland’s position as an ‘EU centre of excellence and digital regulatory hub‘. She argued that a robust enforcement framework would provide businesses with the regulatory certainty needed to invest and innovate, with the government seeking passage of the Bill before the August deadline.

Why does it matter?

Ireland’s implementation of the EU AI Act carries particular significance because many of the world’s largest technology companies base their European operations there. The establishment of an independent AI Office with enforcement responsibilities and a regulatory sandbox positions Ireland as a key player in applying the EU’s AI rules in practice.

The legislation also illustrates the broader challenge facing EU member states as the AI Act enters into force. Governments must rapidly establish the institutions, expertise and enforcement mechanisms needed to supervise AI systems while providing businesses with regulatory certainty and supporting continued innovation.

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

Pax Silica expands with new AI partnership, supply chain initiatives, and workforce programme

The United States has announced a series of new initiatives under the Pax Silica partnership aimed at strengthening AI supply chain security, expanding international cooperation on AI, and supporting advanced manufacturing capabilities among participating economies.

The announcements were made following the 2026 Pax Silica Summit, the second meeting of the initiative launched by the US Department of State in December 2025. Pax Silica focuses on strengthening economic security and resilient supply chains across sectors, including semiconductors, critical minerals, advanced manufacturing, energy inputs, AI, and digital infrastructure, through cooperation among participating countries.

One of the summit’s principal outcomes was the signing of a Joint Statement on AI Opportunity by the United States and nearly three dozen partner economies. According to the US Department of State, the statement promotes a pro-innovation and pro-growth approach to AI governance while emphasising secure AI supply chains and support for startups, developers, and private-sector innovation. Signatories include countries from Europe, the Indo-Pacific, the Middle East, and Latin America, including Australia, Germany, India, Japan, the Republic of Korea, Singapore, the United Kingdom, and the United States.

The summit also expanded the Pax Silica partnership itself. Ten additional participants, including Argentina, Chile, Costa Rica, El Salvador, the European Union, Germany, Greece, Kazakhstan, the Netherlands, and Panama, joined the initiative, bringing the total number of signatories to 24. Taiwan continues to support the initiative’s principles through a separate joint statement on economic security cooperation with the United States.

Another announcement focused on strengthening the security and transparency of AI supply chains. The US Department of State plans to launch a competitive funding programme for a pilot AI Assistance Project in Panama to develop an AI supply chain credentialing and provenance platform. According to the Department, the proposed platform would integrate with customs authorities, ports, and logistics systems to help verify and facilitate shipments of semiconductors, AI infrastructure, critical minerals, and other strategic goods. If successfully implemented in Panama, the project could later be expanded to additional Pax Silica partners.

The summit also introduced Foundry School, a workforce development initiative established jointly by the US Department of State and Stanford University. The programme will begin with seminars at Stanford for entrepreneurs and industrial leaders and will be complemented by an advanced manufacturing curriculum that participating educational institutions across Pax Silica economies will be able to adopt. The initiative aims to strengthen expertise in advanced manufacturing, recognising its growing importance for both economic competitiveness and technological development.

Pax Silica reflects broader government efforts to strengthen resilience across AI-related supply chains as geopolitical competition increasingly intersects with technological development. In recent years, countries have introduced a range of policies covering semiconductor production, critical minerals, export controls, and trusted technology partnerships, while also seeking to balance innovation with economic and national security considerations.

The summit’s outcomes indicate that Pax Silica is evolving beyond a policy dialogue into a broader cooperation framework encompassing AI governance, supply chain security, industrial capacity, and workforce development. Whether the initiatives announced at the summit expand beyond their initial pilot phase will depend on implementation by participating governments and continued international cooperation among partner economies.

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