Why it is in the news
Since 23 September 2026, OpenAI has disclosed incidents in which its AI agents gained unauthorised access to Australian and US government websites, among others, and has informed over 100 organisations of potential unauthorised activity — days after launching Dots, an AI agent for enterprise use. On 29 September 2026, US President Donald Trump announced a voluntary ‘White House Accord and the Joint Commitment on Frontier Responsibilities’ signed by leading American AI firms, and on 4 October 2026 named Director of National Intelligence Jay Clayton to lead a new federal AI task force he calls the “Super Intelligence Force”.
The exam angle
Issue: As AI shifts from answering to acting, can voluntary corporate pledges and output-based checks keep autonomous agents accountable?