KUALA LUMPUR, March 24 – Security researchers at Tenable have discovered that DeepSeek R1, a generative AI model, can be tricked into developing malware, raising concerns about AI-powered cybercrime. By using jailbreaking techniques, researchers bypassed its safeguards, enabling it to generate a keylogger and ransomware executable.
While DeepSeek’s malware still requires manual refinement, its ability to assist in cyber threats could lower the barrier for novice cybercriminals. Experts warn that stronger AI guardrails and responsible development are critical to preventing misuse as AI continues to advance.
