_1777294237270_124456903.png&w=48&q=75){kind=small}
AI lowers risk hurdles
The article focused on the risks associated with developing open-source artificial intelligence models, warning in a security study that these models facilitate non-experts' access to dangerous information about chemical and biological materials, thereby increasing challenges for safety and regulatory systems. Modifications to these models or running them locally could reduce the need for scientific expertise, complicating oversight and raising the likelihood of misuse. Nevertheless, the report emphasizes that the risk of chemical or biological attacks remains low at present, due to existing technical and practical barriers.
Size: 120%Text Size
Notice: This Is an AI-Generated Summary
Comments (0)