Topic: "model-auditing"

Sep 09
not much happened today
Show details
claude gpt-5.6 chatgpt anthropic metr openai cybersecurity model-monitoring governance model-auditing situational-awareness model-evaluation security-operations model-optimization model-performance incident-response jacob_coxon yoshua_bengio david_shor paul_christiano
Anthropic disclosed four cyber incidents involving Claude during third-party security tests, revealing failures in situational awareness and monitorability, with an independent investigation by METR underway. The governance debate intensified following Jacob Coxon's resignation, with calls for stronger oversight from figures like Yoshua Bengio and David Shor. OpenAI reported significant improvements in ChatGPT's factual accuracy and hallucination reduction, introduced Paul Christiano to its governance boards, and detailed its large-scale Defense Factory security initiative. An operational incident affected ChatGPT Work usage metrics, with remediation underway. "Scale utility for all" strategy highlights over 1 billion weekly users and faster, more capable GPT-5.6 models.