Stamos: Open-Weight Models Push Cybersecurity Into 'Dark Age'

Alex Stamos, the newly installed chief information security officer at Cognition, told The Verge's Decoder podcast that cybersecurity has entered what he calls a dark age.

"The metaphor I've been using is that we are entering a dark age," Stamos said. He argued that open-weight models have reached the point where "you can automate most of the kill chain," and that the sequence has flipped: instead of stockpiling exploits, an attacker's AI "just goes hog wild and it looks at the targets and it finds the bugs in real time and it writes the exploits in real time."

Stamos, who previously ran security at Facebook and Yahoo, framed the shift as economic. The next wave, he told host Nilay Patel, "is not going to be from OpenAI and Anthropic. It is going to be from open weight models." The targets he named were not frontier labs. "There's going to be a bunch of hugging face attacks except they're not going to be super high-end French AI companies," he said. "They're going to be midsized, you know, medical supply firms, small insurance companies, school districts, the kind of people who get ransomware all the time."

He compared the current moment to the early commercial internet. "It's got a real frontier feel to it, just like the '90s, when the whole idea of an always-online world was brand-new," Stamos said.

The interview lands alongside a run of AI-attack stories on our cybersecurity feed: Censys recently catalogued hundreds of exposed Artex-style AI hacking tools online, and OpenAI hired Hacktron this month to red-team its third-party integrations.