CVE-2026-54249: VercelAIAdapter trusts client-controlled `providerMetadata` to construct `UploadedFile` — S3/GCS confused deputy via provider metadata injection
Vulnerability summary
CVE-2026-54249 is a medium (CVSS 6.8) Server-Side Request Forgery vulnerability in pydantic pydantic_ai. EPSS estimates a 0.32% chance of exploitation in the next 30 days, higher than 24.4% of scored vulnerabilities.
Not severe (CVSS 6.8, network access); no exploitation reported by the 5 checked sources
Source freshness unavailable
How is this decided?Timeline
- 29-07-2026PublishedVulnerability first disclosed.
- 30-07-2026Last modifiedVulnerability information updated.
Description
Pydantic AI is a Python agent framework for building Generative AI applications. In versions 1.65.0 through 1.105.0, and 2.0.0b1 through 2.0.0b5, a client that submits message history to a Pydantic AI UI adapter (such as the Vercel AI adapter) can reference arbitrary files in the application's model-provider or cloud-storage account. While file URL parts are validated against a scheme allowlist, UploadedFile references — which point to a file by provider file ID or cloud-storage URI (e.g. s3://…, gs://…) — were forwarded without validation. Because the provider resolves an UploadedFile using the server-side identity (IAM role, service account, or provider API key) rather than the client's, an attacker can craft message history to make the server read objects from its own account or other tenants, given a referenceable identifier. Exploitation requires a valid file identifier, which is not always unguessable depending on how the application names objects. This issue has been fixed in versions 1.106.0 and 2.0.0b6.
CVSS metrics
3 scores · 3 sources · 1 version- OSV · PypiEcosystem advisory · calculated6.8MediumCVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:C/C:H/I:N/A:N
- NVDNIST6.8MediumCVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:C/C:H/I:N/A:N
EPSS trend
Probability
0.32%
Chance of exploitation in the next 30 days
Percentile
24%
Versus all other CVEs
30-day change
No change 0.00 pp
From 0.32% on 04-10-2026
Known exploited
Compare KEV catalogsNot in the CISA KEV catalog.
Not in the ENISA CNW catalog.
Not in the CIRCL catalog.
Public exploits
0 foundNo public exploits indexed.
Weaknesses
CWE · root causeAffected systems
Known conditions: network-reachable, no authentication, no user interaction (CVSS vector). Other conditions: not stated in public data.
Pydantic3 products
- pydantic_ai≥ 1.65.0, < 1.105.0 | 2.0.0:beta1 | 2.0.0:beta2 | 2.0.0:beta3 | 2.0.0:beta4 | 2.0.0:beta5
- pydantic-ai≥ 1.65.0, < 1.106.0 | ≥ 2.0.0b1, < 2.0.0b6
- pydantic-ai-slim≥ 2.0.0b1, < 2.0.0b6 | ≥ 1.65.0, < 1.106.0
PyPI2 products
- pydantic-ai≥ 1.65.0, < 1.106.0 | ≥ 2.0.0b1, < 2.0.0b6
- pydantic-ai-slim≥ 1.65.0, < 1.106.0 | ≥ 2.0.0b1, < 2.0.0b6
2 vendors · 5 products
References
8 links- github.com/pydantic/pydantic-ai/security/advisories/GHSA-h7p7-w5gc-xj3wX Refsource CONFIRMWEBVendor Advisory
- pypi.org/project/pydantic-aiPACKAGE