Anthropic to share Mythos cyber flaw findings with global finance watchdog

Startup has declined to release Claude Mythos AI model publicly amid fears it could be used by hackers

Opens the original publisher in a new tab. Full articles belong to their publishers; a subscription may be required.