ITIF Urges AI-Assisted NVD Overhaul to Address Growing Vulnerability Backlog
Regulatory & Policy · 3 sources · Updated 1mo ago · First seen Oct 12, 2026

The Information Technology and Innovation Foundation submitted recommendations to NIST to modernize the National Vulnerability Database (NVD) as AI-driven vulnerability discovery and faster exploitation strain manual processing. ITIF cited a Department of Commerce Inspector General projection that the backlog could grow from 27,000 cases in 2025 to 60,000 by the end of 2026. Its proposals include AI-assisted validation and enrichment, contributions from trusted external providers, better product identification, and clearer responsibilities between NIST and CISA. Human review would remain necessary for consequential or disputed assessments.
The submission comes amid industry attention to NVD processing delays, including Flashpoint’s guidance on managing the slowdown and VulnCheck’s review of first-half 2025 exploitation activity. ITIF urged prioritization based on exploitation evidence, exploit-chain involvement, reachability, and independently verified AI-demonstrated exploitability—not severity scores alone. It also recommended auditable AI decisions, interoperable machine-readable data, independently verified automated remediation, and transparent processing and data-quality metrics. For security teams, the practical implication is to supplement NVD enrichment with exploitation intelligence and asset-specific exposure data rather than treating database completeness or severity ratings as sufficient grounds for remediation decisions.


