AI Safety Institute

The White House has reversed its decision to restrict the UK AI Safety Institute's access to frontier models from Google, Anthropic, and OpenAI. The policy shift follows direct lobbying led by UK AI Minister Kanishka Narayan, after British officials warned that losing access would undermine international AI safety evaluations and cooperation.

A World Bank jobs council is calling for crisis-level employment support measures, warning that workers need stronger protections as artificial intelligence reshapes labour markets. The appeal comes alongside an International Monetary Fund warning that AI could hit jobs like a 'tsunami', with both institutions urging governments to act now on retraining, safety nets and employment policy.

Elgin Community College in Illinois attempted to cancel its contract with surveillance company Flock Safety, but the company refused to let it out of the deal. The case is drawing attention to how Flock's agreements with customers work and the difficulties public institutions face when trying to exit contracts for AI-powered camera and license plate surveillance systems.

Kupor said a newly formed 'Super Intelligence' task force will pressure companies on artificial intelligence safety. The announcement signals a focus on holding AI developers accountable as concerns grow over the risks of advanced systems. Details on the task force's powers and membership have not been reported.

The Brookings Institution argues that AI agents — autonomous systems that act on behalf of users — need coordinated international governance rather than fragmented national rules. The piece frames the debate over who sets standards for accountability, safety and cross-border oversight as increasingly urgent as agentic AI spreads. It adds to a growing policy conversation about how governments should regulate systems that make decisions independently.

A new guide outlines the range of proposals being put forward for regulating artificial intelligence, covering how governments and institutions are weighing rules on safety, accountability and acceptable use. As AI systems spread rapidly across industries, policymakers, companies and researchers continue to debate how strict guardrails should be and who should enforce them, making the regulatory question one of the most contested issues in technology today.

A new commentary examines the question of who holds authority to declare artificial intelligence systems safe. It highlights a gap between the companies developing AI and the regulators, experts and public institutions expected to oversee it, raising concerns about accountability, transparency and conflicts of interest in safety assessments of increasingly powerful models.

Commentary published by Homeland Security Today argues that artificial intelligence is reshaping what it calls the safety economy, the institutional architecture built around security, risk and control. The piece contends that policymakers and the public have been slow to recognise how deeply AI is rewiring that structure. It frames this lack of recognition as a growing governance problem as AI expands across security and public-safety systems.