AI-Generated Browser Ransomware Abuses Chromium API

Researchers say they’ve documented the first case of a frontier AI model — DeepSeek — being used to construct a working ransomware technique that runs entirely inside the browser. The malware combines a browser-malware concept the model treated as unrealistic with a real Chromium API capability, producing a functional attack path on both Windows and Android. This shows AI-assisted malware development moving from theoretical jailbreak output to a demonstrated, cross-platform working exploit chain. Distribution details or in-the-wild use were not included in the report.