Member

Powerful models, like Anthropic’s Mythos, are redefining the cyber-security landscape with potentially catastrophic effects for our critical national infrastructure. Mythos has exposed previously undiscovered flaws in every major operating system and every major web browser, putting at risk our social and economic security. In the past week alone, the Trump Administration, not exactly known for its heavy handedness in regulating the tech sector, imposed an export control directive on Anthropic’s Fable 5 and Mythos 5, citing national security concerns. This means that the Administration felt so concerned about the model’s effects on national security, that it decided no one outside the US should have access to it. In the UK, the National Cyber Security Centre has predicted that it is highly likely that cyber-attackers will be able to use AI to help them to find zero-day vulnerabilities. Powerful models such as Mythos would aid cyber-attackers in exploiting vulnerabilities in defence systems that our experts had not previously been aware of.