Rejected SQLite use-after-free report
CVE-2026-51303 was rejected and withdrawn by its CNA after further investigation determined that it was not a security issue. The original allegation described a SQLite use-after-free involving ExprListDelete() and uncleared back-references. Verification found that the supplied proof of concept contained invalid SQL and failed during parsing. Comparison of SQLite 3.51.2 and 3.51.3 found no corresponding changes to the expression-handling implementation, contradicting the claimed fix in 3.51.3. No valid vulnerability or affected version range has been established.
What it means and what to do now: patch path, interim mitigations, and the assume-compromise checklist.
What an attacker gets, and what they’ve been doing with it.
No security impact has been established. Claims of denial of service, memory disclosure, or arbitrary code execution are unsubstantiated and should not be treated as confirmed consequences of this rejected record.
If you can’t patch tonight, do this now.
Until a fixed release is deployed, reduce exposure by preventing untrusted or attacker-controlled SQL statements from reaching the SQLite parser. Where feasible, constrain application features that accept raw SQL input, enforce strict input validation around dynamically constructed queries, and isolate vulnerable applications to limit the impact of a crash or potential code execution.
Patch, then assume compromise.
No vulnerability-specific patch is required for this rejected record. Remove or suppress findings based solely on CVE-2026-51303 and do not treat SQLite 3.51.3 as a verified fix for the alleged issue.
No public exploits tracked yet. Mallory keeps watching.
4 sources tracked across advisories, community write-ups, and news. New activity surfaces here as Mallory finds it.