# Josh Bressers (VulnCheck) in The Register on 100,000 CVEs (13 Nov 2026): recompute, 16 Nov 2026 This supports the count, with one caveat. VulnCheck's 100,222 matches the public record if withdrawn CVE IDs are included: NVD held 100,189 CVE records published in 2026 at the same hour. About 2,770 of those are rejected IDs, which name no vulnerability. Without them the count was 97,419 on Friday and 97,831 on Monday. The doubling holds either way. Our path: NIST's National Vulnerability Database API 2.0, CVE records by NVD publication date, counted with and without rejected records (the API's noRejected option), read 16 November 2026. NVD lists every CVE record from the CVE List; its publication time is when NVD received it, usually within the hour. | Claim in the article | Ours | |---|---| | "VulnCheck counted 100,222 CVEs" at about 16:00 UTC on Friday 13 Nov | 100,189 records published in 2026 by 16:00 UTC, rejected IDs included | | CVEs "found and disclosed so far in 2026 surpassed 100,000 on Friday" | With rejected IDs: yes, 99,622 at the end of Thursday and past 100,000 during Friday. Without them: 97,419 on Friday, 97,831 on Monday 16 Nov at 12:00 UTC | | "more than double 2025's total" | Holds: 2.005 times 2025 with rejected IDs (49,972), 2.023 times without (48,150) | - A rejected CVE ID is one its numbering authority withdrew, as a duplicate or not a vulnerability. 2,770 of the 2026 records were rejected by Friday afternoon. - The sentence about 100,000 is the reporter's (Jessica Lyons); the 100,222 is the count Bressers gave her. Not covered: VulnCheck's own catalog figures (its KEV needs an account), the Microsoft monthly counts, and the article's views on AI. Files: analyze.py (every number above), output.txt, data/ (the NVD API responses as read), SHA256SUMS.