AI supply chains are susceptible to risks
Traditional tools fail to address AI model vulnerabilities, compliance, and geopolitical risks, hindering safe AI adoption and innovation.
Manage AI model risks proactively
Identify and block malicious code from model components, validate license compliance, and help ensure model provenance before models enter your environment.
Detect threats in AI models and components
Cisco continuously scans public repositories like Hugging Face for malicious code and vulnerabilities within AI model files. By scanning repositories, potential threats can be identified before they can be introduced into your enterprise environment.
Check for license compliance
Detect and block AI models with risky or restrictive open-source software licenses—such as copyleft licenses like GPL—that pose intellectual property (IP) and compliance risks. This helps to ensure legal adherence and avoids inadvertent IP violations.
Maintain proper provenance of AI models
Flag and enforce policies on AI models that originate from geopolitically sensitive regions. Maintain compliance and mitigate potential risks based on potential geopolitical liabilities.
Accelerate enterprise AI adoption from the beginning
Confirm that third-party and open-source models and components used to build your AI applications are secure and compliant from the start of the development process.
Reduce security risks
Block malicious AI models before they enter your enterprise to prevent security incidents and breaches.
Maintain legal compliance
Avoid IP violations with automated detection of restrictive licenses.
Uphold supplier compliance
Enforce policies based on model origin to align with supplier risk management and corporate compliance requirements.
Network-based enforcement
Prevent risky AI models from entering your enterprise infrastructure through multiple enforcement points, powered by Cisco Security Cloud.
Cisco Secure Endpoint
Detect and quarantine risky AI assets directly from the endpoint.
Cisco Email Threat Defense
Prevent delivery of models and components that violate security policies.
Cisco Secure Access
Monitor and block downloads of high-risk AI assets from online repositories.