• Human Rights and the Regulation of AI (UK Parliament Joint Committee on Human Rights) finds that the current legal and regulatory framework is ill-equipped to prevent or respond to the human rights risks posed by AI. It calls for a new AI Bill and a new independent oversight body with powers to hold both developers and deployers of AI to account.

• Data Sovereignty in Practice (World Economic Forum) argues that storing data within national borders is not enough on its own to achieve data sovereignty, and shows how governments, businesses and civil society can build real control over data, models and digital identity while still working together across borders.

• Thematic Brief of the Independent International Scientific Panel on AI – AI Agents, Misalignment and the Risk of Losing Human Control draws lessons from the OpenAI–Hugging Face incident. It calls for stronger safeguards as AI agents advance.

• AI Risk Management Toolkit (UK Department for Science, Innovation and Technology) gives public sector teams practical guidance on identifying and managing AI risks. It covers the whole life of an AI system and includes a workbook and a risk dashboard.

• A Study of Cybersecurity Literature on Open-Source Software and AI (UK Department for Culture, Media and Sport and Department for Science, Innovation and Technology) finds that very few widely downloaded open-source AI models include any security information. It recommends clear definitions and an “AI Bill of Materials” to record where models and data come from.

• AI Action Plan for Justice: One Year On (UK Ministry of Justice) reports progress on using AI across courts, prisons and probation. It adds a new priority on responding to emerging AI risks, including AI-enabled crime.

• The Grid We Need Now (Independent Review of AI Deployment in the Electricity Networks, UK Government) calls for a move to risk-based, AI-supported grid operations by 2035. It also recommends a national hub to help bring AI into the electricity grid.

• Getting It Right: Building AI Infrastructure That Works for Australia – Consultation Paper (Australian Government) seeks public views on national standards for large data centres and frontier AI training in Australia. It covers energy, water, community impact and safety, and submissions close on 9 October 2026.

• Investment Policy Framework for Digital Transformation (OECD) helps governments attract private investment that supports digital transformation. It builds on existing OECD investment standards.

• The Cyber Gap (Council on Foreign Relations) argues that the US is not prepared for China’s growing cyber threat to its critical infrastructure. It calls for better visibility into Chinese hacking, stronger defences and rebuilt government cyber capacity.

• Serving Citizens 2026 (OECD) shows how governments can make everyday public services, such as passports and benefits, simpler and better joined up. It draws on the first OECD Serving Citizens survey.

• Digital Sovereignty and Europe’s Democracy Stack argues that the EU’s digital sovereignty plans lack funded, independent groups to hold its technology investments to account. It proposes a €1 billion endowment fund to support them.

• Frontier AI Risks: A Practical Way Forward (The Alan Turing Institute) argues that debates about long-term existential risk should not delay action on AI risks that can already be measured. It calls for more testing of deployed AI systems in real settings.

• Updated OECD Guidelines on Measuring Trust and Trust Drivers in Public Institutions helps national statistics offices collect comparable data on public trust.

Articles:

• Participation between Promise and Dead End: Legal Limits to Public Participation in Public-Sector AI Development in the EU, argues that EU law currently gives the public almost no real way to take part in designing the AI systems that governments use. It warns that, without stronger legal rules, public participation may become a formality rather than a real safeguard.

• KafkaGPT: On Algorithmic Bureaucracy and Keeping Law’s Promise, warns that AI in public administration can create systems that look lawful but leave people unable to understand or challenge decisions. It argues that core public law values, such as giving reasons and allowing appeals, must be built into the design of these systems from the start.

• Limiting Artificial Intelligence (with the Rule of Law), argues that the rule of law itself sets limits on how governments can use AI, even where no specific AI rules exist. It concludes that AI systems whose decisions cannot be predicted, explained or challenged should not be used by the state.

• Creative or Uncreative Partner: Comparing Humans and AI in Collaborative Creative Tasks, finds that the apparent creative advantage of AI comes mainly from producing longer answers, not more original ideas. In a study of 140 pairs, working with AI made people’s own creative ideas worse, while pairs of humans became more creative together over time.

• Gendering Conversational AI in Human–AI Interaction: The Role of Interaction Context and Sexist Ideology in Shaping User Evaluations, finds that people give AI a gender based on the situation it is used in, not only on the design choices of its makers. In a study of 833 participants in China, AI in serious, authoritative roles tended to be seen as male, and users’ own attitudes about gender shaped how warm, competent and trustworthy they judged the AI to be.

• Theory of Mind in Action: The Instruction Inference Task in Dynamic Human-Agent Collaboration, tests whether AI agents can work out what a person means when their instructions are unclear or incomplete. It finds that some leading AI models, when given a few worked examples, performed about as well as the 52 human participants in the study.

• Self-Reinforcing Barriers to Municipal Digital Government: A System Dynamics Perspective, explains why local governments struggle to go digital even after the barriers have been identified many times. Based on interviews with 67 municipal staff and GovTech providers in Germany, it finds that the barriers feed into and strengthen one another, which keeps digital progress stuck.

• Shared Digital Leadership in Government: A Roles and Competencies Study, argues that digital change in government depends on leadership skills spread across many roles and levels, not only on senior leaders. It shows that gaps and risks appear when people’s job responsibilities do not match their skills, and offers a way for governments to find these gaps, using cyber security as an example.

• Clicking Away Conflict: How Government-Service Digitalization Shapes Labor Unrest, finds that China’s move to put government services online was linked to fewer labour strikes in its cities between 2010 and 2023. It suggests that online services gave workers new ways to raise complaints and get replies from government, so disputes were handled through official channels instead of strikes.

   

Blogs:

• Don’t Believe the Doomers: Europe’s Tech Industry Slams AI Panic, explores how Europe’s tech industry is pushing back against pessimistic narratives about its ability to compete in AI.

• Evaluation Accelerator Fund: Phase 5 Project Summaries, highlights UK government projects using data and evaluation to improve public-sector policies and decision-making.

• The Real AI Economy Is Being Built by Ordinary People, explores how workers and small business owners across Asia and Africa are using AI in practical, everyday ways to solve local problems, automate tasks and expand what they can accomplish.

• What Ukraine’s Digital Resilience Taught Estonia About Building an Agentic AI State, explores lessons from Ukraine’s digital resilience for Estonia’s development of an agentic AI-enabled state.

• Public-Sector AI Agents Need Permission Registries Before They Scale, argues that governments need robust permission and identity systems before AI agents can operate at scale.

• Every Digital Government Wants to Be Singapore and Estonia, But Almost Nobody Can Say What That Takes, examines the institutional and governance foundations behind successful digital government models.

• From e-Residency to Agent Codes: Estonia’s Next Digital Identity Experiment, explores Estonia’s proposal for separate digital identities for AI agents, allowing them to act on behalf of people or organisations within defined permissions while maintaining accountability and human responsibility.

• Creative and AI: A Case of David and Goliath, examines tensions between AI development and the interests of creators and creative industries.

• AI Leaders Support Call to Slow Pace of Frontier Development as Safety Concerns Grow, explores growing calls for greater caution in frontier AI development amid concerns about potential safety risks.

• Can the Finance Sector Oversee AI Innovation While Maintaining Its Rapid Progress?, examines how financial regulators can oversee AI innovation while preserving the sector’s pace of technological development.

• An opinion piece by ElectraLink’s Chief Data Officer explains how UK energy data is being opened up to organisations outside the energy sector.

• WEF: Chief economists see both big AI data-centre benefits and ‘significant’ local pushback yet to come , examines expectations that AI data-centre investment will support economic growth and productivity while facing significant local resistance over energy, water and environmental impacts.

• Governing with Agentic AI: When Machines Act on Governments’ Behalf ,explores the governance, accountability and oversight challenges created when AI agents act on behalf of governments.

• EU AI Act Digital Omnibus 2026: What It Changes, examines proposed changes to the EU AI regulatory framework and their implications for implementation of the AI Act.

• Making sense of the current options for AI regulation in the UK, explores different approaches the UK could take to regulating AI, from sector-specific rules to broader regulatory frameworks.

• An opinion piece argues that Google’s short-lived AI feature for faking satellite images in Google Earth has damaged trust in all images from war zones.

• Funding: AI-supported research using UKRI social science data, UK Research and Innovation (UKRI)

• New ECCC Call for Proposals under the Digital Europe Programme, European Cybersecurity Competence Centre

• Vision for an AI-enabled Clean Energy System: A call for evidence, UK Government

• ADACities Call for Expression of Interest, European Commission

• Call for Papers: Advancing Research on the Ethics of Artificial Intelligence, UNESCO and the AI and Society Institute at École normale supérieure (ENS-PSL)

• Call for Papers: Governing Agentic AI, Cambridge Forum on AI: Law and Governance

• Funding: UKRI Enabling AI adoption across science and engineering, UK Research and Innovation (UKRI)

• AI4Gov-X Cascade Funding Call, AI4GovAccelerate