CVE-2026-72642

The record

The native inference process that Elasticsearch uses to evaluate uploaded machine learning models accepts a model operation that computes a memory address from an offset supplied inside the model, without validating that the offset stays within the bounds of the underlying storage. A user with the privileges required to upload and deploy a trained model can craft a model that reads and writes memory outside the intended allocation. The result is heap corruption that crashes the inference process, and, with sufficient control over the heap layout, could allow arbitrary code execution in the context of that process.

Affected packages

Related vulnerabilities

• CVE-2026-31912 libpcap BPF interpreter detects neither reaching the end of the filter program buffer due… CWE-823
• CVE-2026-49746 Software installed and run as a non-privileged user may conduct improper GPU system calls… CWE-823
• CVE-2026-21734 A web page that contains unusual GPU shader code is loaded into the GPU compiler process… CWE-823
• CVE-2026-34193 Kernel software installed and running inside a Guest/Host VM may post improper commands… CWE-823
• CVE-2026-28764 MediaArea MediaInfoLib LXF element parsing heap-based buffer overflow vulnerability CWE-823
• CVE-2026-42946 A vulnerability exists in the ngx_http_scgi_module and ngx_http_uwsgi_module modules that… CWE-823
• CVE-2026-41907 uuid is for the creation of RFC9562 (formerly RFC4122) UUIDs. CWE-823
• CVE-2025-33215 NVIDIA SNAP-4 Container contains a vulnerability in the VIRTIO-BLK component where a… CWE-823