Calif says Anthropic model helped build macOS exploit (opens in new tab)

Researchers at cybersecurity firm Calif said they used Anthropic's unreleased Claude Mythos Preview model to help find macOS vulnerabilities and build a working exploit against Apple M5 hardware. Calif said the work was AI-assisted but still required human security expertise. Apple said it was reviewing Calif's report to validate the findings. Several reports said patches were underway or expected after validation. It remained unclear whether a released update had already addressed the same i...