AI Science Tracker

Zoom annotation parser heap buffer over-read

Zoom Workplace clients / Zoom annotation feature · AI involvement attributed to Public frontier AI models used through A Security’s AI research workflow (specific models not identified)

Finding and attribution

A wire-provided character count can allocate a glyph buffer larger than the received body; copying the shorter body leaves uninitialized heap bytes that the parser may return/read, and Zoom’s CNA states this can allow denial of service. The vendor bulletin lists distinct affected product branches and fixes; claims beyond its scope are not inferred.

Product: Zoom Workplace clients / Zoom annotation feature. AI system: Public frontier AI models used through A Security’s AI research workflow (specific models not identified). Team / organization: A Security / Zoom.

CVE identifiers: CVE-2026-53414.

AI and human contribution (source account): A Security’s primary account explicitly identifies CVE-2026-53414 as one of the first two annotation vulnerabilities found in its research. It states the finding and exploit work used fewer than 20 prompts to publicly available AI models; the account describes human-led target selection/reverse engineering and documents model-directed analysis, concrete reproduction and exploitation steps. Zoom’s own security bulletin confirms the case-specific vulnerability and credits Idan Levcovich of A Security as reporter. The model names and exact division of work for this CVE are not stated.

Human involvement is only described where the source account establishes it; unknown involvement does not mean none. Vendor confirmation of a flaw does not confirm every claim about AI contribution.

Discovery context: Zoom’s bulletin and CVE record were publicly published 2026-08-11; A Security’s technical disclosure states public disclosure on 2026-08-11. Its account states discovery on June 8, reporting June 10 and fixes in June/July, before public disclosure. No source reports exploitation in the wild.

Remediation status: fix documented. Zoom’s bulletin lists Zoom Workplace, VDI, Rooms, Meeting SDK and Video SDK fixed-version cutoffs; the separate bulletin revision history notes product/version updates. The vendor publishes client-side fixes and server-side mitigation dates. This does not establish installed patch state.

A source-documented fix does not establish installed patches or current safety. Source checking does not independently reproduce the finding or establish observed exploitation.

Disclosed 11 Aug 2026. AI attribution dated 11 Aug 2026. Sources checked 17 Nov 2026; these are separate dates, not a live installation check.

Recorded CVSS: Medium · highest recorded score 6.5. Documented technical effects: denial of service.

Severity, consequences and exploitation evidence

Case grouping uses the highest recorded CVSS score, with its version-specific severity band. It is descriptive, not a combined or editorial risk index. All recorded assessments remain below; scores from different versions and sources may disagree. CVSS specifications define version-specific scoring and severity bands.

  • CVE-2026-53414: CVSS 3.1 · 6.5 (Medium) · CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:N/I:N/A:H. Zoom Security Bulletin · checked 17 Nov 2026.

CVE-2026-53414: denial of service.
Prerequisites / affected scope: An attacker must participate in a meeting and send malformed annotation data to an affected Zoom client. Do not claim data disclosure: although A Security reports reading heap content in a controlled test, the Zoom CNA assigns only denial of service in its impact description..
Exploitation evidence: The researcher reports a controlled demonstration; Zoom’s bulletin describes a possible denial of service. No exploitation in the wild is reported.. Impact source · checked 17 Nov 2026.

Documented fix: Zoom’s bulletin lists Zoom Workplace, VDI, Rooms, Meeting SDK and Video SDK fixed-version cutoffs; the separate bulletin revision history notes product/version updates. The vendor publishes client-side fixes and server-side mitigation dates. This does not establish installed patch state.

Original evidence

Download this source record (JSON)

Open this case in the interactive tracker · Browse all static security cases