VulnerabilityPublished 29 Jul 2026

CVE-2026-54249: pydantic-ai · Server-Side Request Forgery (SSRF)

Affects pydantic-aiSource CVE Database V5

Not known to be exploited · CVSS 6.8 Medium : upgrade to 2.0.0b6 or 1.106.0 when convenient.

Data as of 27 Oct 2026, 17:11 UTC

CVSS 3.1 base score

6.8Medium

How severe the flaw is on the 0–10 CVSS scale.

EPSS: 0.3% chance of exploitation in 30 days – higher than 24% of CVEsExploit Prediction Scoring System: the likelihood attackers will use this flaw soon.
Patch: Available – fixed in 2.0.0b6, 1.106.0

CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:C/C:H/I:N/A:N

Attack Vector: Network
Attack Complexity: High
Privileges Required: None
User Interaction: None
Scope: Changed
Confidentiality: High
Integrity: None
Availability: None

Bold values are the worst case: easiest to exploit or largest impact.

Attack Vector: Network: The vulnerable component is bound to the network stack and the set of possible attackers extends beyond the other options listed, up to and including the entire Internet. Such a vulnerability is often termed 'remotely exploitable' and can be thought of as an attack being exploitable at the protocol level one or more network hops away (e.g., across one or more routers).
Attack Complexity: High: A successful attack depends on conditions beyond the attacker's control. That is, a successful attack cannot be accomplished at will, but requires the attacker to invest in some measurable amount of effort in preparation or execution against the vulnerable component before a successful attack can be expected. For example, a successful attack may require an attacker to: gather knowledge about the environment in which the vulnerable target/component exists; prepare the target environment to improve exploit reliability; or inject themselves into the logical network path between the target and the resource requested by the victim in order to read and/or modify network communications (e.g., a man in the middle attack).
Privileges Required: None: The attacker is unauthorized prior to attack, and therefore does not require any access to settings or files to carry out an attack.
User Interaction: None: The vulnerable system can be exploited without any interaction from any user.
Scope: Changed: An exploited vulnerability can affect resources beyond the security scope managed by the security authority of the vulnerable component. In this case, the vulnerable component and the impacted component are different and managed by different security authorities.
Confidentiality: High: There is total loss of confidentiality, resulting in all resources within the impacted component being divulged to the attacker. Alternatively, access to only some restricted information is obtained, but the disclosed information presents a direct, serious impact.
Integrity: None: There is no loss of integrity within the impacted component.
Availability: None: There is no impact to availability within the impacted component.

Definitions: FIRST CVSS v3.1 specification.

Exploitation evidence

Disclosed: evidence found.
CVE published 29 Jul 2026
Detectable: no evidence on record.
No public nuclei template indexed
Weaponised: no evidence on record.
No public exploit code indexed in Metasploit or Exploit-DB
Exploited in the wildExploited: no evidence on record.
Not in CISA KEV and no source reports exploitation
Observed by OffSeq MirageSeen by Mirage: no evidence on record.
No attack traffic recorded by OffSeq Mirage sensors

A missing step is not evidence of safety. Sources: cve.org (opens in a new tab), the Metasploit, Exploit-DB and nuclei-templates index as of 26 Oct 2026.

Disclosed: The CVE record is public on cve.org. Its publication is t = 0.
Detectable: A public nuclei template can find exposed instances, for defenders and attackers alike.
Weaponised: Public exploit code exists in Metasploit or Exploit-DB, or CISA SSVC or the CVSS exploit maturity reports a public proof of concept.
Exploited in the wild: CISA KEV lists it, CISA SSVC rates exploitation active, or a source reports attacks.
Observed by OffSeq Mirage: OffSeq Mirage honeypot sensors recorded attack traffic for this CVE, or found exposed hosts without attack traffic.

Moments are spaced evenly, not to scale; a dashed line marks a gap of six months or more. Deltas count from t = 0. Sources that give only a date (CISA KEV, Exploit-DB, or a time of exactly 00:00Z) read “same day” rather than hours. Metasploit modules carry the flaw’s disclosure date, not their release date, and nuclei templates carry no date, so both appear above but not on the timeline.

What to do

Upgrade pydantic-ai and pydantic-ai-slim to 2.0.0b6 or 1.106.0.

Affected software

Product	Affected	Fixed in
pydanticpydantic-ai	Affected >=2.0.0b1 <2.0.0b6	Fixed in 2.0.0b6
pydanticpydantic-ai	Affected >=1.65.0 <1.106.0	Fixed in 1.106.0
pydanticpydantic-ai-slim	Affected >=2.0.0b1 <2.0.0b6	Fixed in 2.0.0b6
pydanticpydantic-ai-slim	Affected >=1.65.0 <1.106.0	Fixed in 1.106.0
PyPIpydantic-ai`pkg:pypi/pydantic-ai`Other threats	Affected >=1.65.0 <=1.105.0	Fixed in–

Check your own hosts for these packages with threat-finder (opens in a new tab), our free open-source scanner.

AI-suggested

A fix is available in pydantic-ai versions 1.106.0 and 2.0.0b6, which validate UploadedFile references similarly to file URLs. Users should upgrade to these versions to remediate the vulnerability. If upgrading is not feasible, users should avoid passing untrusted client-submitted message history to the agent or strip UploadedFile parts from incoming messages before processing to prevent exploitation.

Description

### Summary A client that can submit message history to a Pydantic AI UI adapter can reference arbitrary files in the application's model-provider or cloud-storage account. The server forwards the reference to the model provider, which fetches it using the server's own credentials, allowing the client to read files it should not have access to. ### Details UI adapters reconstruct file parts from client-submitted message history and forward them to the model provider. File **URL** parts are validated against a scheme allowlist before being forwarded, but `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, a client that crafts message history containing an attacker-chosen `UploadedFile` can cause the server to read objects belonging to its own account or to other tenants, given a referenceable identifier. ### Impact Applications that pass untrusted client-submitted message history to an agent through a UI adapter (such as the Vercel AI adapter). Exploitation requires the attacker to reference a valid file identifier; depending on how the application names objects, such identifiers are not always unguessable. ### Patches Upgrade to `1.106.0` (1.x) or `2.0.0b6` (the 2.x beta line), which validate `UploadedFile` references on client-submitted messages the same way file URLs are validated. ### Workarounds If users cannot upgrade, do not pass untrusted client-submitted message history to the agent, or strip `UploadedFile` parts from incoming messages before running the agent.

Impact

An attacker can exploit this vulnerability to read files that the server has access to but the attacker should not, by submitting crafted message history containing UploadedFile references. This can lead to unauthorized disclosure of sensitive data stored in the server's model-provider or cloud-storage accounts. The CVSS score is 6.8 (medium severity), reflecting high confidentiality impact but no integrity or availability impact. Exploitation requires knowledge or guessability of valid file identifiers.

Summary

The vulnerability in pydantic-ai arises from improper validation of UploadedFile references in client-submitted message history forwarded to model providers. While file URLs are validated against a scheme allowlist, UploadedFile references (e.g., provider file IDs or cloud storage URIs like s3:// or gs://) are not validated, allowing an attacker to cause the server to fetch arbitrary files using its own credentials. This can lead to unauthorized reading of files belonging to the server's account or other tenants if valid file identifiers are guessed or known. The vulnerability affects versions >=1.65.0 <=1.105.0. Patches in 1.106.0 and 2.0.0b6 introduce validation for these references to prevent exploitation.

CWE-918Server-Side Request Forgery (SSRF) (opens in a new tab)
Base weaknessCWE Top 25 (2025): #22OWASP Top 10 A01:2025: Broken Access Control (opens in a new tab)
The web server receives a URL or similar request from an upstream component and retrieves the contents of this URL, but it does not sufficiently ensure that the request is being sent to the expected destination.
OWASP Cheat Sheets
Server Side Request Forgery Prevention (opens in a new tab)
Unvalidated Redirects and Forwards (opens in a new tab)
Cross Site Scripting Prevention (opens in a new tab)
DOM based XSS Prevention (opens in a new tab)
Authorization (opens in a new tab)
OWASP ASVS 4.0.3
V12.6.1 (opens in a new tab)
V5.2.6 (opens in a new tab)
Attack patterns
Server Side Request Forgery (opens in a new tab)
From MITRE CWE, MITRE CAPEC and OWASP (Top 10, OpenCRE).

References

github.com/pydantic/pydantic-ai/security/advisories/GHSA-h7p7-w5gc-xj3w (opens in a new tab)

Community reviews

Loading reviews…

Exploitation evidence

What to do

Affected software

Description

AI analysisMachine-generated, updated 13 Aug 2026, 18:11 UTC

WeaknessesCWE-918

Technical details

References

Community reviews

Related threats