Cyber security companies turn AI from threat to opportunity

The release of Anthropic’s new AI model, Mythos, has triggered concerns around how the technology could disrupt the cyber security industry. 

At the end of March, it was reported that AI lab Anthropic had built a new general purpose model that was a “step change” in abilities. The company confirmed that this month when it claimed the model was too powerful to release publicly because it had identified security flaws in “every major operating system and every major web browser”. 

Instead, Anthropic is making a limited release to major technology companies so they can secure their systems ahead of time. Anthropic has called this Project Glasswing. The 10 companies being given access include cyber security giants CrowdStrike (US:CRWD) and Palo Alto Networks (US:PANW), as well as Apple (US:AAPL), Microsoft (US:MSFT) and Nvidia (US:NVDA).