Medium severity · published Oct 8, 2026
CVE-2026-107293
Pydantic AI OpenTelemetry instrumentation: retry prompt content is not redacted when `include_content=False`
5.3
CVSS
0.3%
EPSS
No
On KEV
Description
Pydantic AI is a Python agent framework for building applications and workflows with Generative AI. From 0.3.4 until 1.107.4 and 2.27.1, OpenTelemetry instrumentation configured with InstrumentationSettings(include_content=False) can export retry prompts outside tool calls in gen_ai.input.messages and pydantic_ai.all_messages. Agents using NativeOutput, PromptedOutput, or output validators on text output can therefore disclose validation feedback, including invalid model values quoted by that feedback, to readers of the telemetry backend. Tool-call retries and deployments that do not use include_content=False are not affected by this specific path. This issue is fixed in versions 1.107.4 and 2.27.1.
CVSS vector
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:N
Attack vector
Network
Complexity
Low
Privileges
None
User interaction
None
Confidentiality
Low
Integrity
None
Availability
None
Scope
Unchanged
Exploitation
Is it being used?
Exploit prediction (EPSS)0.3% · higher than 22% of CVEs
Exploited in the wildNo evidence yet
CISA KEVNot listed
Ransomware campaign useNot known
Affected products
Which versions
PY
pydantic-ai · pydantic
>= 0.3.4, < 1.107.4; >= 2.0.0b1, < 2.27.1 · source: vendor
PY
pydantic-ai-slim · pydantic
>= 0.3.4, < 1.107.4; >= 2.0.0b1, < 2.27.1 · source: vendor
PA
Pydantic AI · Pydantic
0.3.4 before 1.107.4; 2.0.0 before 2.27.1 · source: NVD
References
NVD
https://nvd.nist.gov/vuln/detail/CVE-2026-107293
CVE.org
https://www.cve.org/CVERecord?id=CVE-2026-107293
x_refsource_CONFIRM
https://github.com/pydantic/pydantic-ai/security/advisories/GHSA-3gh4-cghq-f8v4
x_refsource_MISC
https://github.com/pydantic/pydantic-ai/pull/7357
x_refsource_MISC
https://github.com/pydantic/pydantic-ai/commit/fe9dbed7b7ccf7e7128b5786886e4441f6f5594f
x_refsource_MISC
https://github.com/pydantic/pydantic-ai/releases/tag/v2.27.1
More from pydantic-ai
8.6
CVE-2026-25580
Pydantic AI Affected by Server-Side Request Forgery (SSRF) in URL Download Handling
7.6
CVE-2026-107295
`pydantic-ai-slim` web UI `/api/chat` accepts browser-simple cross-origin requests that can trigger agent tool execution
7.5
CVE-2026-107286
Pydantic AI: Concurrency-limited models can keep their slot when a streamed request ends early
6.8
CVE-2026-107289
Pydantic AI: SSRF cloud-metadata blocklist bypass via IPv6 zone identifier (incomplete fix for CVE-2026-46678 and CVE-2026-48782)
6.8
CVE-2026-54249
VercelAIAdapter trusts client-controlled `providerMetadata` to construct `UploadedFile` — S3/GCS confused deputy via provider metadata injection