AI is reshaping cybersecurity as automated attacks accelerate

Cybersecurity is entering a period of rapid change as artificial intelligence helps attackers discover vulnerabilities and exploit them in real time, according to Alex Stamos, chief security officer at Cognition and a longtime security specialist. He described recent weeks as a succession of breaches affecting AI companies, neocloud providers and other technology firms.

Stamos said attacks are increasingly being chained together. In one example, stolen API keys from Fireworks allegedly enabled access to Cursor; a malicious plugin was then distributed to 5,000 companies, creating the possibility of thousands of further breaches. He also cited the Binance hack as the largest heist in history, while noting that different criminal and state-backed groups are involved in current attacks.

He said open-weight AI models can now automate much of the cyber “kill chain”: examining targets, locating bugs, writing exploits and deploying them. He pointed to ARTEX, a Chinese open-source tool, as one system being discussed, and said its capabilities mark a return to the “Wild West” atmosphere of the 1990s. In his view, the cybersecurity industry built around skilled researchers, responsible disclosure and bug bounties is being challenged because powerful tools are becoming accessible without the former technical hierarchy or professional culture.