AI TRENDS | Malicious Code Found in Mistral AI Software Package Distributed via PyPI

Microsoft has reported that attackers have inserted malicious code into a Mistral AI software package distributed through PyPI, leading to malware execution on Linux systems. According to NS3.AI, Microsoft indicated that the malware has the capability to steal credentials. Mistral, however, stated that there is no evidence suggesting that its infrastructure was compromised. The company further noted that the incident is connected to a larger supply-chain attack associated with the TanStack security breach.