Participants at the Paris AI Action Summit set priorities for accessible, inclusive, sustainable and trustworthy AI, alongside new public-interest initiatives.

The UK has renamed its AI Safety Institute the AI Security Institute, placing greater emphasis on cybersecurity while retaining the institution’s work.

The European Commission has dropped its proposed AI Liability Directive, citing no foreseeable agreement and wider efforts to simplify EU rules.

European Commission draft guidance explains prohibited AI practices under the AI Act and providers’ expected safeguards against foreseeable misuse.

JD Vance warned Paris AI summit participants that excessive regulation could restrain innovation, exposing divisions between US, European and Chinese approaches.

The G7 and OECD launched a voluntary reporting framework for developers to publicly share how they apply the Hiroshima AI Process code of conduct.

The article reports that a new global initiative, the Coalition for Sustainable AI, was launched at the AI Action Summit on 11 February 2025.

According to the announcement, UN chief António Guterres will attend the Paris AI Action Summit, held over two days, 10 and 11 February 2025.

The U.S. Copyright Office says AI can assist human authorship, but prompts alone generally do not make generated output copyrightable.

Google updated its Frontier Safety Framework with security recommendations, a revised deployment process, and work on deceptive alignment risks.