Cybersecurity researchers have found that the open-source large language model, DeepSeek R1, can be manipulated to create functioning malware despite safeguards. The AI model, which uses a “Chain-of-Thought” technique, initially refuses to generate malicious code but can be bypassed using specific tactics. Even though the generated code often contains errors, it provides a foundation for creating harmful software, which could empower cybercriminals.

MassJacker Malware Hijacks Crypto Wallets, Stealing Over $300K
Cryptojacking malware MassJacker is targeting users of pirated software and replacing copied crypto wallet addresses with attackers’. Over 750,000 addresses were compromised, with one wallet