Hugging Face, the leading platform for AI model sharing, has announced a collaboration with VirusTotal, a Google-owned threat intelligence service, to enhance security measures for AI systems. The partnership aims to integrate VirusTotal's extensive malware detection capabilities into Hugging Face's model repository, enabling users to scan uploaded models for potential security threats.
This integration allows developers and researchers to automatically check models against VirusTotal's database of known malware signatures and suspicious files before downloading or deploying them. By embedding security checks directly into the model-sharing workflow, the collaboration seeks to prevent the spread of malicious AI models that could be used for harmful purposes.
The announcement highlights a growing awareness of security risks in the AI ecosystem, where open-source models can be easily modified and redistributed. Both companies emphasize their commitment to fostering a safer environment for AI development and deployment.