CVE-2021-41228: TensorFlow saved_model_cli code injection through eval on user strings

Summary

TensorFlow's `saved_model_cli` tool calls `eval` on user-supplied strings, which allows code injection. An attacker can run arbitrary code on the platform where the CLI runs, though the maintainers note the impact is limited because the tool is always run manually.

Mitigation

Patched by adding a `safe` flag that defaults to `True`, plus an explicit warning for users. The fix will be included in TensorFlow 2.7.0 and cherry-picked to TensorFlow 2.6.1, 2.5.2 and 2.4.4.

Affected packages in the Exposure Registry

Matched by package name and ecosystem. Each entry shows whether the package delegates to a language model and how many tracked packages depend on it.

• tensorflowPyPINo LLM dependency detected · 26 tracked dependents
• tensorflow-gpuPyPINo LLM dependency detected · 2 tracked dependents

Related items

• MediumCVE-2026-121936: easyVoice server-side request forgery through openaiBaseUrl in /api/v1/ttsSimilar attack · NVD/CVE Database
• MediumCVE-2026-118628: MLflow get-artifacts server-side request forgery information disclosureSimilar attack · NVD/CVE Database
• HighCVE-2026-118625: MLflow Tracking Server privilege escalation via get_endpointsSimilar attack · NVD/CVE Database
• InfoWhen trust becomes the payload: Inside the threats, etc quarterly reportSimilar attack · Elastic Security Labs
• MediumCritical Langflow flaws enable authentication bypass and RCESimilar attack · CSO Online