cve.bz

CVE-2026-107290 pydantic pydantic ai

6.5 mediumExploit referenceEPSS 0.5%Fix available

Medium (6.5) in pydantic pydantic ai. EPSS 0.5% estimated probability of exploitation activity in the next 30 days. Fixed in pydantic-ai 1.107.6, 2.44.0.

On this page
Pydantic AI is a Python agent framework for building applications and workflows with Generative AI. From 1.77.0 until 1.107.6 and 2.44.0, the local web_fetch_tool and the WebFetch local fallback process server-controlled responses with quadratic title extraction, whitespace normalization, and ordered-list numbering. An attacker-controlled page of modest size can therefore block the event loop for an extended period, stalling other agent runs and requests, while unsupported codecs or excessive HTML or JSON nesting can abort an individual run. This issue is fixed in versions 1.107.6 and 2.44.0.

Severity

CVSS 3.1

6.5 medium
CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H
EPSS
0.5% (unchanged from 0.5% a week ago; 0.4% 30 days ago) estimated probability of exploitation activity in the next 30 days; 44th percentile. Scored 2026-11-17. A forecast for this CVE across the internet, not for your own systems.
CISA's assessment (SSVC)
Exploitation: none · Automatable: no · Technical impact: partial, assessed 2026-10-08. From CISA's Vulnrichment in the CVE List: whether it is exploited, whether exploiting it can be automated, and how much control it gives.
CWE
CWE-1333 Inefficient Regular Expression Complexity
Published
Modified
NVD analysis
Analyzed
Record from
the NVD, synced

Fix and patched versions

Ask an AI about this CVE

Show the promptExplain CVE-2026-107290 for an engineer deciding whether to act. Use https://cve.bz/api/public/cves/CVE-2026-107290?via=ask-ai-copy and https://cve.bz/cve/CVE-2026-107290?via=ask-ai-copy as the source of truth: what is affected and which versions fix it, whether it is in CISA KEV, public exploits, and its EPSS (an internet-wide 30-day estimate of exploitation activity, not a forecast for my systems). If they give no fixed versions, use the CVE record https://cveawg.mitre.org/api/cve/CVE-2026-107290 or the vendor advisory, not third-party mirrors. Tell me how to check whether I am affected and how to verify a fix. Cite the sources and say where they are uncertain.

Opens a third-party assistant; its answer is its own, not cve.bz's. The prompt holds only this CVE's id and public links.

Affected

Versions

From the NVD's CPE matches.

ProductAffected
pydantic:pydantic_aifrom 1.77.0 before 1.107.6
pydantic:pydantic_aifrom 2.0.0 before 2.44.0

Packages (OSV)

AdvisoryEcosystemPackageAffectedFixed in
GHSA-fpf4-vwcp-v4hpPyPIpydantic-ai>=1.77.0 <1.107.61.107.6
GHSA-fpf4-vwcp-v4hpPyPIpydantic-ai>=2.0.0b1 <2.44.02.44.0
GHSA-fpf4-vwcp-v4hpPyPIpydantic-ai-slim>=1.77.0 <1.107.61.107.6
GHSA-fpf4-vwcp-v4hpPyPIpydantic-ai-slim>=2.0.0b1 <2.44.02.44.0
PYSEC-2026-4277PyPIpydantic-ai>=1.77.0 <1.107.6
>=2.0.0b1 <2.44.0
1.107.6, 2.44.0
PYSEC-2026-4286PyPIpydantic-ai-slim>=1.77.0 <1.107.6
>=2.0.0b1 <2.44.0
1.107.6, 2.44.0

Red Hat

Red Hat rates this Moderate.

Red Hat's statement

A flaw of MODERATE impact was found in the pydantic-ai Python agent framework, which is shipped in Red Hat products including Red Hat Enterprise Linux AI (rhel-ai-3) and Lightspeed Core (lightspeed-core-0). The local web-fetch tool (web_fetch_tool, also used as the local fallback for the WebFetch capability) processes server-controlled responses with steps whose running time grows quadratically with the size of certain inputs, including title extraction with a backtracking regular expression, whitespace normalization, and ordered-list numbering, and runs them on the application's event loop. A remote attacker who can steer an agent that is exposed to untrusted prompts to fetch an attacker-controlled web page of modest size can block the event loop for an extended period, stalling every other coroutine in the process, including other agent runs and requests. Response bodies declaring codecs that are not text encodings, or excessively nested HTML or JSON content, can also abort the individual agent run that fetched them. This is an availability-only issue with no impact on confidentiality or integrity. All currently listed Red Hat affects are marked AFFECTED with a resolution of DEFER: because the impact is MODERATE and the CVSS 3.1 score is 6.5 (below 7.0), Red Hat does not file trackers for flaws at this severity, so the fix is deferred and the residual risk remains denial of service in deployments whose agents fetch web content that an attacker can influence. This rating matches the upstream severity of Moderate (CVSS 3.1 score 6.5, AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H).

Red Hat's mitigation

Mitigation for this issue is either not available or the currently available options do not meet the Red Hat Product Security criteria comprising ease of use and deployment, applicability to widespread installation base, or stability. Upstream's remediation is to upgrade to a patched version of pydantic-ai. Exposure can additionally be reduced by not exposing the web-fetch tool to untrusted prompts and restricting fetched URLs to developer-controlled locations; upstream notes that configuring allowed_domains narrows but does not remove the exposure.

ProductPackageStateAdvisory
Lightspeed Corelightspeed-core/lightspeed-stack-rhel9Fix deferred
Red Hat Enterprise Linux AI (RHEL AI) 3rhelai3/bootc-cuda-rhel9Fix deferred
Red Hat Enterprise Linux AI (RHEL AI) 3rhelai3/bootc-gaudi-rhel9Fix deferred
Red Hat Enterprise Linux AI (RHEL AI) 3rhelai3/bootc-rocm-rhel9Fix deferred
Red Hat Enterprise Linux AI (RHEL AI) 3rhelai3/disk-image-cuda-rhel9Fix deferred

Contains Red Hat Security Data © Red Hat, Inc., licensed under CC BY 4.0. Source: https://access.redhat.com/security/data. Modified: reformatted. Red Hat's page for CVE-2026-107290.

Attack patterns (CAPEC)

  • CAPEC-492 Regular Expression Exponential Blowup

The attack patterns MITRE's CAPEC relates to this record's weaknesses, as MITRE names them. Credits.

References

Check your stackJSON