Cisco launches Model Provenance Kit for AI lineage | Let's Data Science
· Source: letsdatascience.com
In short
Cisco has released an open-source Model Provenance Kit, a Python toolkit and CLI for tracing AI model lineage, combining metadata, tokenizer signals, and weight-level fingerprints. The toolkit includes pairwise comparison, database scanning, streaming support for models over 20 GB, and aggregates eight provenance signals into a single pipeline score. Aimed at reducing supply-chain, vulnerability, licensing, and regulatory risks tied to unverified third-party models, the release includes a reference database of fingerprints for approximately 150 base models across 45+ model families and 20+ publishers.
Read Original Article — Read the full article on the publisher’s site.
In the issue of week 19, 2026: AI-Driven Platforms Boost Resiliency and Efficiency Across Global Supply Chains
Tags: AI lineage, supply-chain, provenance, open-source, security, model risk