Anthropic's Mythos sparks security fears

The release of Mythos has set the global tech industry on edge, as the model demonstrates advanced capabilities in autonomous code writing and cybersecurity exploitation.

Anthropic has opted for a controlled release to trusted companies, citing the risk of the model being used to crack sensitive cryptography.

This move has prompted a response from rival OpenAI, which launched its ownGPT-5.4-Cyber model designed for defensive workflows.

The competition underscores the growing tension between rapid AI advancement and the necessity for robust national security safeguards.