// Security Advisory Triage

PyJWT 2.14.0 patches high-severity BOM bypass vulnerability

GitHub advisories — pip · published 2026-09-30 21:59 UTC
ACTION NEEDED PATCH_NOW SECURITYAPI
WHAT CHANGEDGitHub published CVE-2026-102272 on 29 September 2026, a high-severity BOM bypass in PyJWT. The vulnerability is patched in version 2.14.0.
WHY IT MATTERSPyJWT handles JSON Web Token validation in Python applications. A BOM (byte order mark) bypass can allow an attacker to craft tokens that pass validation checks they should fail, potentially bypassing authentication or authorization controls.
WHAT TO DOAudit your dependency trees for PyJWT versions below 2.14.0 and upgrade to 2.14.0 or later. If you cannot upgrade immediately, review all JWT validation code paths for exposure to untrusted token input.
Model-reported confidence 85% (not measured accuracy) Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
+ghsa_id: GHSA-r6x4-923q-g947
+summary: PyJWT BOM Bypass
+severity: high
+published_at: 2026-09-29T23:14:00Z
+cve_id: CVE-2026-102272
+vulnerabilities[].package.name: PyJWT
+vulnerabilities[].first_patched_version: 2.14.0
HOW IT WAS JUDGED
Triage signals below are model-reported assertions, preserved verbatim; they are not independent arithmetic verification.
triage verdictMATERIAL at 0.92, needed 0.5
triage reasoningPyJWT is a standard Python dependency for JWT handling. High-severity BOM bypass with patch available warrants immediate assessment.
signalsPyJWT is a widely-used Python package in dependency trees · high severity rating · new CVE with patched version identified · BOM bypass suggests authentication/validation bypass · first_patched_version present indicates actionable fix
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)