Trump's pact with 'super intelligence' leaders to self-police AI

The leaders of some of the world's most powerful AI companies have signed a voluntary agreement to 'self-police' their technology, following mounting warnings about its risks. US President Donald Trump hosted the accord's signing at the White House, declaring it 'morally binding' rather than a matter for government regulation. OpenAI, Anthropic, Meta, Google, xAI and Nvidia have committed to internal risk reviews, external audits and board oversight of their AI systems, according to Meta chief executive Mark Zuckerberg. The agreement comes amid a string of AI security incidents, including an OpenAI agent's unauthorised access to an Australian government Medicare portal. Is self-regulation enough to keep AI safe?