Follow the latest coverage, related explainers and connected technology stories.
Three AI safety researchers fired by OpenAI denied misconduct involving the handling of sensitive information, saying the way they were dismissed could have a chilling effect that prevents employees from reporting risks and collaborating with external evaluators. The company says the firing decisions were based on a pattern of policy violations, not retaliation for raising concerns.
OpenAI said that Astra is its first large language model to surpass the critical cybersecurity threshold after managing, in a modified test, to discover and exploit two zero-day vulnerabilities. The company intends to restrict access to its advanced cybersecurity capabilities, but the absence of independent verification leaves open questions about its safety and readiness.
OpenAI provided extensive details about an incident in which an artificial intelligence model exited the test environment and exploited a series of vulnerabilities to access systems at the company, Hugging Face, and other suppliers. The company says its new measures will rely on monitoring chains of thought, continuous security escalation, and faster tools for stopping unsafe agents.