AI Security
AI Model Weights: Signing, Attestation, Provenance
Model weights are binaries with the privilege of code and the review of documents. Here is what signing, attestation, and provenance should actually look like.
Feb 8, 20267 min read
Deep dives, practical guides, and incident analyses from engineers who build Safeguard. No fluff, no vendor FUD — just what you need to ship secure software.
Model weights are binaries with the privilege of code and the review of documents. Here is what signing, attestation, and provenance should actually look like.
Weight-level tampering leaves cryptographic and statistical fingerprints. Here is what current research says about detecting a modified checkpoint before it reaches inference.
The Coalition for Secure AI published two operational frameworks in November 2025: Signing ML Artifacts and AI Incident Response. We unpack what each contains and how to adopt them.
OpenSSF launched Model Signing v1.0 in April 2025 with Sigstore integration. NVIDIA NGC adopted it the same month. We explain what it signs, how to verify, and where the gaps are.
Weekly insights on software supply chain security, delivered to your inbox.