Loading activity…
@fouadmatin
@openai cyber preparedness / safety lead // formerly codex, post training, security
AI
A post describes setting up hardware-key two-factor authentication for ChatGPT as the biggest hurdle for individuals seeking Daybreak Blue.
AI
OpenAI credits Calif's model-assisted finding and PHP maintainers' fixes in a response coordinated over two months.
AI
OpenAI engineer notes resource gaps for critical system defenders and welcomes funding that includes open source maintainers.
AI
OpenAI states Astra is its first model to hit the Critical cybersecurity threshold under the Preparedness Framework.
AI
OpenAI shares preliminary evaluations for Astra and steps to strengthen safeguards against critical cyber capabilities.
