Anthropic withholds Mythos AI model after tests by JPMorgan, Apple, Google and more

 AI firm Anthropic has developed a powerful tool, Claude Mythos Preview, capable of finding complex software flaws. Due to its potential misuse, the company is withholding public access, instead partnering with tech giants like Apple, Google, and Microsoft through Project Glasswing. This initiative aims to leverage the AI’s capabilities to identify and patch critical vulnerabilities, bolstering global cybersecurity efforts. AI firm Anthropic has developed a powerful tool, Claude Mythos Preview, capable of finding complex software flaws. Due to its potential misuse, the company is withholding public access, instead partnering with tech giants like Apple, Google, and Microsoft through Project Glasswing. This initiative aims to leverage the AI’s capabilities to identify and patch critical vulnerabilities, bolstering global cybersecurity efforts.