After Claude Mythos, OpenAI developing cybersecurity AI may pilot privately
Technology
OpenAI is working on a fresh AI model aimed at making cybersecurity smarter, especially after Anthropic's Claude Mythos wowed everyone by spotting thousands of security flaws that humans missed.
Instead of a big public launch, OpenAI may first share the tool with a handful of companies.
Spud 2 years in development
Codenamed "Spud," this project has been two years in the making.
OpenAI President Greg Brockman hinted it might be a big step toward super-smart AI (AGI), and Codex head Thibault Sottiaux believes Spud could even rival Claude Mythos.
showing OpenAI's focus on sharing powerful tech responsibly.