GPAC CVE-2026-109827

Exploitation requires GPAC (≤26.07.0) to parse an attacker-controlled MPEG-2 transport stream whose AIT (Application Information Table) section declares more application descriptors than the fixed application_descriptors_id array can hold, triggering the out-of-bounds write when index_app_desc_id exceeds the array bounds. The key LIMITING factor is user interaction (UI:P in the vector): the vulnerable code path only runs when someone - or an automated workflow - feeds the malicious stream into GPAC, so there is no unsolicited remote trigger. No authentication or special privileges are needed (PR:N), and no non-default configuration is required beyond invoking the M2TS/AIT parsing path; realistic risk is therefore concentrated in environments that automatically ingest untrusted media (transcoding services, upload processors).

This is a medium-severity (CVSS 4.0 base 5.3) out-of-bounds write in GPAC's MPEG-2 TS AIT parser, and the signals converge on a moderate-to-low real-world priority rather than an emergency. The CVSS 4.0 vector (AV:N/AC:L/AT:N/PR:N/UI:P) indicates no authentication is required and low attack complexity, but it explicitly requires user interaction (UI:P) - the victim must process an attacker-supplied media stream - which caps the impact metrics at Low across confidentiality, integrity, and availability (VC:L/VI:L/VA:L), with no scope change. The exploit maturity is marked E:P (proof-of-concept), consistent with the publicly posted POC archive; this means publicly available exploit code exists but is distinct from confirmed active exploitation. The CVE is NOT in CISA KEV and no EPSS probability was supplied in the input, so widespread opportunistic exploitation cannot be inferred. SSVC-style reasoning: exploitation is at 'proof-of-concept,' it is not automatable (requires a user to open a crafted file), and technical impact is partial (memory corruption bounded to Low per the vendor scoring). Net: a legitimate bug to patch on a normal cycle, not a drop-everything issue, unless GPAC is used in an automated/server-side media-processing pipeline that ingests untrusted streams.

An attacker crafts a malicious MPEG-2 transport stream containing an AIT (Application Information Table) with more application descriptors than the fixed-size descriptor array can hold, then delivers it via a web download, email attachment, or streaming URL. When a user or an automated pipeline processes the file with a vulnerable GPAC build (≤26.07.0), gf_m2ts_decode_ait writes past the bounds of application_descriptors_id, corrupting adjacent heap memory. A public proof-of-concept (poc_gpac_g1.zip) has been released, lowering the barrier to reproduction.