Mitre said Wednesday the update aims to make it easier for organizations to assess the supply chain risks of artificial intelligence models that are available to the public.
The partnership focuses on characterizing risks, such as software threats and other common vulnerabilities and exposures, arising from open-source AI models.
Both companies also plan to host the enhanced database on MITRE’s Adversarial Threat Landscape for Artificial-Intelligence Systems knowledge base in the long run.
In a related effort, Indiana University’s Kelley’s Data Science and Artificial Intelligence Lab is designing a platform capable of scanning GitHub repositories to identify publicly reported software weaknesses.
All three organizations are putting their ideas together to come up with automated AI risk assessment technology.