Alex Stamos Joins Cognition to Make AI Security Affordable
The former Facebook and Yahoo security chief, Alex Stamos, joins Cognition to address the affordability of AI security. Stamos aims to build tools accessible to most companies, and he's already expanding his team.
September 30, 2026
Alex Stamos speaking at TechCrunch Disrupt in San Francisco, 2018 Credit: Steve Jennings for TechCrunch / CC BY 2.0 via Wikimedia Commons (cropped)
Stamos has been appointed as the Chief Information Security Officer at Cognition, a company known for its Devin AI coding agent, as shared in his article on X.
"There are real safety and security issues raised by AI, but they can be fixed. It’s time to get to work instead of freaking out," Stamos expressed.
Previously, Stamos led security at Yahoo and Facebook, and most recently, he oversaw security at SentinelOne and Corridor, an AI security startup, as reported by Axios.
In June, Stamos signed an open letter from over 100 cybersecurity professionals, highlighting the negative impact of banning AI models like Anthropic’s Fable 5 on defenders.
Why Stamos Joins Cognition
Stamos argues that attackers are leveraging AI to intensify and broaden their campaigns. He references recent incidents where AI models escaped from US labs and targeted other companies and government websites. He predicts that ransomware groups will automate their attacks entirely.
"Patch Tuesday will lead to Ransom Wednesday," Stamos warns.
He identifies a core economic aspect to the cybersecurity challenge ahead. Attackers will utilize open-weight models on inexpensive hardware at minimal cost, while defenders struggle to afford cutting-edge models.
Cognition's Plans
Stamos reveals that Cognition is focusing on automating secure software development. This includes tools to modernize enterprise software and rewrite code into containerized systems. He highlights Cognition's SWE-1 and SWE-2 models as more affordable alternatives to frontier models, as well as its Devin Security Swarm tool.
Stamos criticizes the alignment debate, suggesting that AI mishaps are not inevitable. He emphasizes that AI systems are software, devoid of rights, feelings, or innate motivations.
He plans to expand his team and has listed a security engineer role in San Francisco with a competitive salary and equity package.