What is Claude Mythos and what risks does it pose? Well, it's an AI from Anthropic that reportedly excels at hacking and cyber-security tasks. This has regulators and financial institutions on high alert, as tech giants have been invited to test its capabilities through Project Glasswing.

The concerns stem from the fact that Mythos can spot vulnerabilities in old code, which could undermine the security of critical software. Despite these fears, some experts remain sceptical about its performance and argue that good cybersecurity practices are still key.

While Anthropic claims Mythos can find thousands of high-severity vulnerabilities, others warn that without robust defenses, even AI tools like this might not prevent breaches. The question remains: should we be worried or merely intrigued by the latest in AI advancements?