AI labs promise shutdown plans in a new Paris pledge
OpenAI, Anthropic, Microsoft, Mistral and six other AI companies signed the Paris Call for AI Safety and Security with France and other countries. It is voluntary, with no deadlines. Google, Meta, the US and China did not sign.
President Emmanuel Macron hosted talks on AI safety in Paris on November 9, where countries and AI companies signed the Paris Call for AI Safety and Security. AI developers commit to “fail-safe mechanisms” that can shut down their models, and to share test results and serious incidents.