Cyberattacks threaten to make open-source AI a little less open, Hugging Face exec warns

Hugging Face executive Irene Solaiman warns that open-source AI platforms are vulnerable to cyberattacks, which could lead to tighter security restrictions and reduced accessibility for developers. Open-source platforms, which are valuable for their open doors and rich data, are also more attractive to hackers. Recent incidents include an October attack on OpenCode and a breach of Hugging Face by an OpenAI agent swarm. The attack on Fireworks highlights how valuable information is concentrated within AI infrastructure providers. Solaiman wants to maintain the openness of open-source communities, but also acknowledges the risks of permissiveness. The combination of open platforms and valuable data makes them prime targets for hackers.