Model Supply Chain
A concept in enterprise AI security focused on the security risks that arise when organizations adopt and deploy AI models. Key concerns include the integrity of pre-trained model provenance, contamination of fine-tuning data, access control for inference services, and the potential misuse of model permissions throughout the deployment lifecycle.
Timeline (last 90 days)
Model supply chain risks include the origin and integrity of pre-trained models, fine-tuning data poisoning, inference service access control, and privilege abuse from models being used in automated pipelines.
Incorporating AI models into existing asset management and threat modeling frameworks is the most actionable starting point for security teams.
All Facts (2)
Model supply chain risks include the origin and integrity of pre-trained models, fine-tuning data poisoning, inference service access control, and privilege abuse from models being used in automated pipelines.
50%UnverifiedIncorporating AI models into existing asset management and threat modeling frameworks is the most actionable starting point for security teams.
50%