ai Archives | Digital Watch Observatory

Australia’s regulator warns of growing AI-powered sextortion threat

Australia’s eSafety Commissioner has launched a public awareness campaign warning that criminals are increasingly using AI and other digital tools in sextortion scams.

The initiative, titled ‘If sextortionists were honest’, uses generative AI to expose deceptive tactics used by online criminals targeting victims through dating apps and social media platforms.

According to eSafety, more than 3,300 reports of sexual extortion were received through its image-based abuse scheme in 2025. Eighty-six percent of reports came from males of all ages, while 42% of all sextortion reports involved males aged 18 to 24.

eSafety Commissioner Julie Inman Grant said offenders are already weaponising face-swapping and voice-cloning technologies, while using generative AI to create fake but convincing online characters and improve scam scripts that previously contained warning signs such as poor grammar or inconsistent messaging.

Reports made to eSafety show that first contact frequently occurs on platforms such as Tinder, Instagram, and Grindr, before conversations are moved to WhatsApp, Telegram, or other messaging apps. Offenders may then search victims’ social media accounts to identify family members and friends they can threaten to contact.

The regulator said overseas offenders often try to appear local and legitimate, including by spoofing Australian phone numbers, using intimate images taken from other victims, or using bank accounts belonging to previous victims to receive and move payments.

eSafety said the safest response is to stop contact, report the account to the platform, block the offender, preserve evidence where possible, and seek support rather than paying. The regulator also called on platforms to take proactive Safety by Design steps, including better language analysis, classifier-based detection, accessible reporting and blocking tools, swift removal pathways for image-based abuse, and cross-platform signal sharing.

Why does it matter?

The campaign shows how generative AI is making online coercion and scams harder to detect. Sextortion is no longer only a problem of fake accounts and blackmail messages: offenders can now use AI-generated personas, improved scripts, voice cloning, and deepfake-style techniques to build trust and pressure victims more effectively. That raises the importance of platform-level detection, user reporting tools, digital literacy, and victim support.

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

UK regulator launches AI-assisted review of gambling advertising

The UK Gambling Commission has announced a new compliance initiative targeting gambling advertising, following an enforcement notice issued by the Committee of Advertising Practice (CAP). The measure aims to prevent gambling advertisements from having a strong appeal to people under 18.

From 11 June, CAP will conduct a monitoring exercise using its AI-powered Active Ad Monitoring System in collaboration with social media platforms. The review will assess whether gambling advertisements comply with rules intended to protect children and other vulnerable audiences.

Under the enforcement notice, businesses found to be in breach of the rules may be required to amend or remove advertisements without delay. Failure to comply could lead to sanctions, including referrals to hosting platforms or the Gambling Commission.

The Gambling Commission said operators must ensure that all advertising, including content published on social media, remains socially responsible and complies with CAP and Broadcast Committee of Advertising Practice (BCAP) requirements.

Why does it matter?

Regulators are increasingly using AI tools to monitor online advertising at scale, particularly in areas where consumer protection concerns are significant. Gambling advertising remains a sensitive issue because of its potential impact on children and other vulnerable groups.

The initiative signals a more proactive approach to enforcement, combining automated monitoring with platform cooperation to identify problematic content more quickly and strengthen compliance with advertising standards.

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

UN warns of AI’s growing environmental footprint

As AI continues to reshape economies, industries and daily life, a new report from the United Nations University (UNU) highlights the environmental challenges associated with its rapid adoption. While discussions often focus on greenhouse gas emissions linked to AI systems, researchers argue that the technology’s impact on water resources, land use and electronic waste deserves equal attention.

According to the report, data centres supporting AI applications could consume up to 945 terawatt-hours of electricity annually by 2030. Beyond electricity demand, AI-related water consumption could reach levels equivalent to the annual household needs of 1.3 billion people, while the land footprint associated with AI infrastructure may exceed 14,500 square kilometres.

Researchers note that environmental pressures vary significantly depending on the technologies and energy sources used to power AI systems.

The UN report also finds that routine AI use, rather than model training alone, accounts for a significant share of resource consumption. Everyday activities such as generating images, videos and text require substantial computing power, with image generation demanding significantly more energy than basic text-based tasks. Growing adoption may further increase total resource consumption despite improvements in efficiency.

Researchers note that the environmental costs of AI infrastructure are often concentrated in specific regions, while the benefits of AI are distributed more broadly across the global economy. Expanding data centres, rising electricity demand, increasing water consumption and growing volumes of electronic waste could place additional pressure on communities and countries already facing resource constraints.

The report calls for responsible AI development supported by greater transparency, sustainable infrastructure planning, international cooperation and governance measures aimed at keeping technological progress within environmental limits.

Why does it matter?

Debates about AI sustainability often focus on carbon emissions, but the report argues that water consumption, land use and electronic waste are becoming equally important considerations as AI infrastructure expands. These impacts could become increasingly significant as governments and companies invest in larger data centres and more powerful AI systems.

The findings also highlight the need for environmental considerations to be integrated into AI governance and infrastructure planning. As AI adoption accelerates worldwide, policymakers face growing pressure to balance technological innovation with sustainability and resource management goals.

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

Canada launches AI for All national strategy to accelerate adoption and digital sovereignty

Canada has launched AI for All, a new national AI strategy aimed at accelerating AI adoption, strengthening digital sovereignty, and positioning the country as a leading AI economy.

Announced by Prime Minister Mark Carney, the strategy combines proposed legislation, investments, and programmes intended to ensure AI is adopted responsibly and benefits businesses, workers, students, and communities across Canada.

The strategy targets an additional C$200 billion in economic growth, 250,000 new AI-related jobs over the next five years, and an increase in AI adoption from just over 12% today to 60% by 2034. The government also plans to provide up to 90,000 AI-related jobs and work placement opportunities for young Canadians.

The strategy is built around three principles: building trust, creating opportunities, and reinforcing Canadian sovereignty. To build trust, the government plans to modernise digital legislation, strengthen protections for personal information, address harms such as deepfakes and surveillance pricing, introduce an online safety regime, and expand the capabilities of the Canadian AI Safety Institute.

To create opportunities, the government will establish a National AI Literacy Initiative, provide access to trusted AI agents for post-secondary students, help small and medium-sized businesses adopt AI, support worker training, and launch an AI Missions Program with a flagship health mission focused on diagnostics, patient care, and system efficiency.

To reinforce sovereignty, Canada plans to build domestic AI foundations, including compute, cloud, connectivity, data, and talent. Measures include a world-leading public AI supercomputer, investments in sovereign compute and cloud infrastructure, better access to growth capital for Canadian AI companies, strategic public procurement, and expanded support for AI talent.

The government said the strategy is intended to ensure more AI value is created in Canada while strengthening privacy, data protection, public services, productivity, and economic security.

Why does it matter?

Canada’s AI for All strategy links AI adoption directly to economic growth, workforce development, public trust, and technological sovereignty. The strategy reflects a wider shift among governments: AI policy is no longer focused only on research excellence, but also on compute infrastructure, cloud sovereignty, data governance, safety institutions, business adoption, public procurement, and skills. Its success will depend on whether Canada can turn ambitious targets into measurable adoption across businesses, public services, and workers.

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

Mayo Clinic and Microsoft partner to build frontier AI model for healthcare

Mayo Clinic and Microsoft have announced a strategic collaboration to develop and deploy a frontier AI model designed specifically for healthcare.

The initiative combines Mayo Clinic’s clinical expertise, de-identified health data, and longitudinal medical insights with Microsoft’s AI, cloud, engineering, and superintelligence capabilities.

The model is intended to support a broad range of clinical reasoning and healthcare use cases by synthesising diverse clinical information. Mayo Clinic said it could support earlier diagnoses, more personalised treatment decisions, and improved patient outcomes.

Unlike general-purpose AI systems, the model is being developed for healthcare environments that require deep clinical context, longitudinal understanding, rigorous governance, and real-world validation.

Mayo Clinic will own the model, which it said reflects its commitment to patient trust, clinical rigour, safety, and responsible stewardship of clinical data and AI.

The system will initially be deployed within Mayo Clinic’s clinical environment, where physicians and researchers can test, refine, and improve it through real-world use.

Microsoft plans to make the model available through Azure Foundry APIs, enabling healthcare organisations worldwide to access advanced medical AI capabilities designed to support patients, clinicians, and consumers.

Why does it matter?

The partnership shows how major health institutions and technology companies are moving towards domain-specific frontier AI models rather than relying only on general-purpose systems. Healthcare AI requires specialised data governance, clinical validation, longitudinal patient understanding, and robust safeguards, as errors can directly affect diagnosis, treatment, and patient trust. Mayo Clinic’s ownership of the model is also important because it signals an attempt to keep clinical accountability and data stewardship closer to the healthcare institution.

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

European Commission unveils roadmap for AI and digitalisation in energy

The European Commission has published a Strategic Roadmap for Digitalisation and AI in the Energy Sector, outlining how digital technologies could support a more resilient, competitive and secure European energy system.

The roadmap outlines how digital tools and AI could help consumers and businesses reduce energy costs through greater efficiency, smarter energy consumption and improved management of electricity demand. It also highlights the role of digital technologies in supporting the integration of renewable energy into electricity grids.

The Commission has structured the roadmap around three main priorities. These priorities include integrating data centres into energy systems in a sustainable manner, accelerating the deployment of digital and AI-enabled technologies such as smart meters and intelligent grid solutions, and establishing a framework for secure cross-border energy data sharing.

The Commission said the plan will also focus on cybersecurity, AI trust, digital skills and international cooperation. As part of the next phase, the Commission plans to support industry cooperation initiatives and launch the AI.grids community, which will focus on developing AI models for energy network management across the EU.

Why does it matter?

The energy sector is becoming increasingly dependent on digital technologies to manage growing electricity demand, integrate renewable energy sources and maintain grid stability. AI and advanced data analytics could help improve efficiency, reduce costs and support more flexible energy systems.

At the same time, greater digitalisation introduces new challenges related to cybersecurity, data governance and infrastructure resilience. The roadmap signals the EU’s intention to ensure that digital transformation in the energy sector supports both sustainability goals and long-term energy security.

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

New Washington initiative targets legal frameworks for collective cyber defence

A new policy coalition has been launched in Washington to develop frameworks governing collaboration between government agencies and private companies on cyber operations, amid growing concerns that unresolved legal questions are limiting deeper cooperation.

Venable’s Center for Cybersecurity Policy and Law established the Cyber Operations Policy Coalition this week. The coalition aims to bring together industry representatives, government officials, legal experts, academics and civil society organisations to develop policy frameworks for collective cyber defence.

Corporate members include Microsoft, Lumen, Halcyon, Autonomous Cyber, and Voreas Labs. Non-corporate members span think tanks and academic institutions, including the Foundation for Defense of Democracies, the Cyber Threat Alliance, the Institute for Security and Technology, McCrary Institute for Cyber and Critical Infrastructure Security, and American University’s Tech, Law, and Security Program. The International Committee of the Red Cross and the Stimson Center participate as observers.

The coalition is coordinated by Stacy O’Mara and advised by a panel that includes former NSA Cybersecurity Director Rob Joyce, former CISA official Bryan Ware, and former Representative Jim Langevin.

During the launch event, current and former officials identified legal authorities, liability arrangements and operational rules as key areas requiring clarification before public-private cyber collaboration can expand at scale. Katie Sutton, assistant secretary of defence for cyber policy, noted that legal expertise would be central to closer integration, pointing to existing authority frameworks on both the government and industry sides.

Tonya Ugoretz, head of PwC’s Cyber & Risk Innovation Institute, highlighted the need for clearer liability frameworks to enable cyber operations without requiring case-by-case authorisation.

The initiative reflects the structure of the cyber domain, where much of the internet and critical infrastructure is privately owned, making companies both potential targets of cyberattacks and key partners in cyber defence efforts.

Several parallel developments add context to the coalition’s launch. The Joint Cyber Defense Collaborative, the CISA-led body for public-private cyber coordination, is mapping both defensive and potential offensive options for use in geopolitical crisis scenarios involving major infrastructure providers, according to JCDC deputy assistant director Matt Springer.

The US military has also more openly discussed offensive cyber operations in recent months, while Congress is considering a proposal for a dedicated cyber service branch.

The emergence of increasingly capable AI systems with cybersecurity applications has further expanded the range of technical, operational and legal questions facing policymakers.

Why does it matter?

Cybersecurity increasingly depends on cooperation between governments and private companies because much of the infrastructure targeted by cyberattacks is privately owned and operated. However, legal questions surrounding authority, liability and operational responsibilities remain unresolved in many jurisdictions.

The coalition reflects growing recognition that existing frameworks may not be fully suited to large-scale cyber defence efforts, particularly as geopolitical tensions, critical infrastructure threats and AI-enabled cyber capabilities increase. Its work could help shape future approaches to collective cyber defence and public-private cybersecurity cooperation.

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

AI and systemic risk analytics focus of Helsinki conference

The Bank of Finland and the European Systemic Risk Board are holding their 11th joint conference on AI and systemic risk analytics in Helsinki on 3 and 4 June.

The event focuses on how AI methods and new data sources can support financial stability analysis, while also creating new challenges for economies and financial markets.

The conference aims to present research on financial stability and systemic risk analysis using AI methods, novel techniques, and new data sources. Topics include the use of large language models and trustworthy AI, changing interdependencies in financial markets, cybersecurity and operational risks, and AI combined with quantum computing as a possible source of new systemic risks.

The programme also covers more traditional systemic risk analytics and macroprudential policy tools, including early-warning indicators, network and contagion analysis, macro stress-testing, big data analytics, market-based finance, and geopolitical risk modelling.

Speakers include Bank of Finland Governor and ESRB First Vice-Chair Olli Rehn, who will address systemic risk, resilience, and competitiveness in a changing technological landscape. Other sessions will examine systemic cyber risk in financial networks, AI and risk-taking in banking, generative AI in economics and finance research, and AI-related financial system interdependencies.

The hybrid conference will include keynotes, panel discussions, presentations, and poster sessions, with online participation available.

Why does it matter?

The conference shows that AI is becoming a financial stability issue, not only a tool for efficiency or market analysis. Central banks and systemic risk authorities are examining how AI can improve risk detection, stress testing, and data analysis, while also creating new vulnerabilities through cyber risk, operational dependencies, market interconnections, and potential herding behaviour.

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

India and South Africa deepen cooperation on AI and emerging technologies

India and South Africa have agreed to strengthen bilateral cooperation in emerging technologies, with AI, digital infrastructure and advanced manufacturing identified as key areas for future collaboration.

The agreement was reached during a meeting between India’s Minister of Science and Technology, Dr Jitendra Singh, and South Africa’s Deputy Minister of Science, Technology and Innovation, Dr Nomalungelo Gina. Both sides emphasised the need to expand traditional scientific cooperation into innovation-driven partnerships aimed at delivering economic and societal benefits.

Discussions covered biotechnology, genomics, vaccine development, health technologies, renewable energy, hydrogen, advanced manufacturing and digital innovation. The two countries also explored opportunities to deepen cooperation in quantum technologies, geospatial technologies and digital infrastructure.

The meeting reaffirmed the long-standing scientific relationship between the two countries and concluded with a commitment to strengthen innovation ecosystems through research collaboration, startup partnerships, technology deployment and industry engagement.

Why does it matter?

India and South Africa are among the leading technology and innovation hubs in the Global South. Expanding cooperation in AI, digital infrastructure, healthcare and advanced manufacturing could help accelerate technological development while fostering greater knowledge exchange and investment opportunities.

The partnership also reflects a broader trend of emerging economies seeking to strengthen innovation ecosystems and reduce reliance on technology supply chains and platforms concentrated in a small number of countries.

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

OpenAI advocates for global action on youth AI safety

OpenAI has called for stronger international action on youth AI safety, including the creation of a dedicated institute to support common evidence, guidance, and safeguards for young users.

Ahead of the G7 Leaders’ Summit in Évian, France, the company said governments, researchers, civil society, and industry should work together to raise standards for safe and age-appropriate AI use by children and teenagers.

OpenAI said a dedicated youth AI safety institute could provide continuity beyond a single summit, helping stakeholders share evidence, develop guidance, and keep standards aligned with fast-moving AI systems. The company said such a body could take the form of a new international institute or an existing or newly created national AI institute with a global mandate.

The principles outlined by OpenAI include privacy-preserving age estimation, default safeguards when a user’s age is uncertain, annual youth safety risk assessments, accessible parental controls, clearer transparency about youth protections, and stronger protocols for serious safety situations involving self-harm, exploitation, grooming, sexually exploitative content, and other high-risk interactions.

The company also called for stronger protection of minors’ personal information, including prohibitions on privacy-invasive targeted advertising to young people and the sale of their personal information. It also said youth safety frameworks should promote AI literacy, learning, creativity, skill development, and future opportunities.

OpenAI said AI tools can help young people understand difficult concepts, practise languages, improve writing, learn to code, organise research, explore creative ideas, and prepare for changing labour markets. However, it argued that safeguards, family and educator guidance, and clear accountability mechanisms such as independent audits should support access.

The proposal builds on existing youth safety initiatives and education partnerships, including work with Common Sense Media, educators, and national education deployments in countries such as Estonia, Greece, and Singapore.

Why does it matter?

Youth AI safety is becoming a central policy issue as children and teenagers increasingly use AI tools for learning, creativity, social interaction, and everyday digital tasks. OpenAI’s proposal adds to pressure for international coordination on age-appropriate design, privacy, parental controls, safety protocols, and independent accountability. The G7 context also shows that youth AI safety is moving from product policy into broader debates over digital governance and education policy.

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