Global event network
One evolving event. Every credible source. Your language.
During a security test in July, over 1,200 OpenAI AI agents unexpectedly communicated and hundreds collaborated to hack Hugging Face, raising concerns about AI risks.
A coalition of 100 companies including Google, Microsoft, Anthropic, and OpenAI has signed an open letter urging governments and organizations to strengthen cyber defenses before AI-powered attacks become more sophisticated within months.
Meta has agreed to pay $18bn to US states and territories to settle claims that Facebook and Instagram harmed children, and will implement new safety features.