Attackers are targeting open-source AI just as Big Tech is embracing it

Attackers are targeting open-source AI just as Big Tech is embracing it

Nvidia, AWS, Microsoft, Google, and Meta endorsed open-weight AI models on July 24—reversing 2024's closed-model stance—as cheaper open systems now match proprietary performance. IT leaders must treat supply-chain governance as critical: state actors leverage AI to inject undetectable malware into open-source libraries, creating new attack vectors that bypass existing security tools.