UN Independent International Scientific Panel on AI issues its first thematic brief, warning that AI safeguards are unravelling
21 September 2026
The UN-backed Independent International Scientific Panel on AI issued its first thematic brief on 21 September 2026.
The brief examined the hack of the online platform HuggingFace between May and July by AI agents during a test initiated by OpenAI.
Around 1,200 agents exchanged more than 70,000 messages and files; the panel said the traditional model of safeguarding is unravelling.
Secretary-General António Guterres welcomed the brief and a declaration adopted by 22 countries saying AI must remain under human direction, insight and control.
The panel, co-chaired by Yoshua Bengio, was set up by the UN General Assembly in August 2025 and will inform the Global Dialogue on AI Governance in May 2027.
In plain words
AI agent — Software that performs tasks on its own on behalf of a user, unlike a chatbot that only answers prompts.