CVE-2026-107291 · GHSA-4x9p-g9wm-8q7f

Pydantic AI OpenTelemetry instrumentation: exception events on tool and agent run spans include content when `include_content=False`

Risk score: 44/100
CVSS: 4.3 v3.1
Exploit chance (30 days): 0.5%
Published: 8 Oct 2026

Open in the live forecast Today's forecast

What to know

What it is: Pydantic AI is a Python agent framework for building applications and workflows with Generative AI. From 0.3.4 until 1.107.6 and 2.44.0, OpenTelemetry instrumentation configured with InstrumentationSettings(include_content=False) can still export sensitive agent content through exception.message and exception.stacktra…
Affected: pydantic-ai (pip) >= 0.3.4, < 1.107.6; pydantic-ai (pip) >= 2.0.0b1, < 2.44.0; pydantic-ai-slim (pip) >= 0.3.4, < 1.107.6; pydantic-ai-slim (pip) >= 2.0.0b1, < 2.44.0
Fix: Upgrade pydantic-ai to 1.107.6 or 2.44.0; pydantic-ai-slim to 1.107.6 or 2.44.0 (whichever matches your release line) or later.
Exploitation: Public proof of concept

Story map

Story map: the flaws, the fix, how it unfolded and what it connects to. Boxes link to the record or article.

At a glance

pydantic-ai: Amber, patch this week. Pydantic AI is a Python agent framework for building applications and workflows with Generative AI. From 2.10.0 until 2.53.0, streamed requests made through ConcurrencyLimitedModel or limit_model_concurrency can retain shared concurrency slots because anyio.CapacityLimiter associates an acquired slot with the borrowin…

The flaws

CVE-2026-107291: Pydantic AI OpenTelemetry instrumentation: exception events on tool and agent run spans include content when `include_content=False` (CVSS 4.3 · EPSS 1%)
CVE-2026-107286: Pydantic AI: Concurrency-limited models can keep their slot when a streamed request ends early (CVSS 7.5 · EPSS 0%)
CVE-2026-107295: Pydantic AI Web chat UI (`Agent.to_web()`, `clai web`): a website visited by the developer can trigger agent runs and server-side tool execution on the local chat endpoint (CVSS 7.6 · EPSS 0%)
CVE-2026-107290: Pydantic AI: Event loop blocked by quadratic title extraction in `web_fetch` (CVSS 6.5 · EPSS 0%)

The fix

Upgrade pydantic-ai: 1.107.6 · 2.44.0 · 2.53.0 · 1.107.4 · 2.28.0 or later on your release line

Who is exposed: pydantic-ai (pip) >= 2.10.0, < 2.53.0; pydantic-ai-slim (pip) >= 2.10.0, < 2.53.0

Where it stands

Public proof of concept

Connects to

Same weakness: CWE-212 (CWE-212). 4 other warnings in 30 days
Same product: Earlier pydantic-ai warnings. 6 in the last 90 days

Why it is Yellow

CVSS 4.3 medium
Public exploit or proof of concept referenced
On your watchlist: PyPI

Affected packages

Package	Ecosystem	Vulnerable	Fixed in
pydantic-ai	pip	>= 0.3.4, < 1.107.6	1.107.6
pydantic-ai	pip	>= 2.0.0b1, < 2.44.0	2.44.0
pydantic-ai-slim	pip	>= 0.3.4, < 1.107.6	1.107.6
pydantic-ai-slim	pip	>= 2.0.0b1, < 2.44.0	2.44.0

References

Original description

Pydantic AI is a Python agent framework for building applications and workflows with Generative AI. From 0.3.4 until 1.107.6 and 2.44.0, OpenTelemetry instrumentation configured with InstrumentationSettings(include_content=False) can still export sensitive agent content through exception.message and exception.stacktrace events, error status descriptions, and model_request_parameters containing instructions or the prompted_output_template. The exposed data is available to readers of the configured telemetry backend and can include tool feedback, provider error bodies, runtime instructions, and structured-output templates even though message attributes are redacted. This issue does not grant new access to agent data, and deployments that do not use include_content=False are not affected by the setting bypass. This issue is fixed in versions 1.107.6 and 2.44.0.

Last activity 8 Oct 2026. Patch Forecast updates every hour.