Check Point says DeepSeek-generated code could become browser-only ransomware

The Register reports that Check Point researchers analyzed a DeepSeek-attributed sample that attempted to build browser-native ransomware and data theft into a web application. The sample was incomplete, but researchers said it could be turned into a working attack with little effort and that DeepSeek V4 could reproduce similar functionality with careful prompting. The case shows how LLM-generated malware ideas can combine with legitimate browser permissions and social engineering.

Key details: Check Point analyzed a DeepSeek-attributed browser-ransomware sample, Researchers said the incomplete sample could be made functional with little effort, The attack abuses browser file-access permissions and social engineering.

Why it matters: This is a concrete example of AI lowering the skill floor for attack development by translating known platform risks into plausible malware workflows.