- NCSC’s Richard Horne says hacking AI tools like Mythos Preview can strengthen defenses if guardrails are in place
- Anthropic’s Mythos Preview, part of Project Glasswing, finds zero-days at scale
- Horne claims that frontier AI quickly uncovers weak fundamentals, giving defenders a chance to decisively outperform cybercriminals
With proper safeguards and safeguards in place, hacking AI tools such as Mythos Preview can be a net positive for cybersecurity defenses everywhere, says Richard Horne, head of the National Cyber Security Center (NCSC).
According to the BBC, Horne repeated these statements in a speech to the NCSC’s annual conference CyberUK on Wednesday.
“As we’ve seen in the media in recent days, frontier AI enables the rapid discovery and exploitation of existing vulnerabilities at scale, illustrating how quickly it will reveal where the fundamentals of cyber security still need to be addressed,” he said.
The article continues below
This time is different
Earlier this month, Anthropic announced a new security initiative called Project Glasswing, and at the center of it is the latest AI model, the Mythos Preview. This model was apparently so good at detecting and exploiting zero-day vulnerabilities that Anthropic decided to only provide it to a handful of large software companies. That way, these companies can get a head start on bad actors before the model is released to the general public at a (yet undetermined) later date.
There’s also plenty of chatter online that this is just a PR stunt, with some people saying that OpenAI did the same for GPT-2, which later turned out to be much more benign.
But this time it really could be different. The Mozilla Foundation said that with the help of Mythos, it managed to find 271 vulnerabilities in Firefox 150, the famous browser’s latest build. When it tried a similar thing with an earlier model – Opus 4.6 against Firefox 148 – it found “only” 22 errors.
Announcing the results, Mozilla CTO Bobby Holley couldn’t hide his excitement, hinting that the cat-and-mouse battle against cybercriminals might finally be coming to an end.
“Our work is not done, but we have turned the corner and can see a future much better than just watching,” he wrote. “Defenders finally have a chance to win, decisively.”
Via BBC
The best antivirus for all budgets
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews and opinions in your feeds.
