/

CVE-2026-40344

Severity:

HIGH RISK

(8.2)

Authentication Bypass in MinIO Object Storage

CVE-2026-40344

Analyzed

Vulnerability type

Category:

Authentication Bypass

CWE:

CWE-287, CWE-306

Impact:

Unauthorized Data Write / Integrity Compromise

Severity:

HIGH RISK

(8.2)

CVSS version:

3.1

Catch this CVE while code is being written

Detect vulnerable code directly in pull requests, so CVEs are fixed before merge.

Summary

An authentication bypass in MinIO's Snowball auto-extract handler (PutObjectExtractHandler) allows attackers who know a valid access key to upload arbitrary objects to any bucket without the corresponding secret key or a valid signature.

By sending specially crafted streaming unsigned-trailer PUT requests, an attacker can bypass signature verification and have tar payloads automatically extracted into target buckets where the access key has WRITE permission.

Affected context

This vulnerability affects MinIO object storage deployments using versions starting in RELEASE.2023-05-18T00-05-36Z and prior to RELEASE.2026-04-11T03-20-12Z.

It specifically occurs in the Snowball-style auto-extract upload path handled by PutObjectExtractHandler when processing S3-compatible PUT requests that use the STREAMING-UNSIGNED-PAYLOAD-TRAILER content mode along with the X-Amz-Meta-Snowball-Auto-Extract: true header.

Any environment where access keys are known, reused, or shared (including the default minioadmin credentials) is at heightened risk because the handler checks IAM permissions but omits cryptographic signature validation for this auth type.

Exploit preconditions

  • Attack surface: MinIO S3-compatible API endpoint exposed over the network (internal or external).
  • Required access:
  • Knowledge of a valid MinIO access key with WRITE permissions for at least one bucket.
  • No need for the corresponding secret key or a correct cryptographic signature.
  • Trigger point:
  • A PUT request to a bucket/object with:
  • X-Amz-Content-Sha256: STREAMING-UNSIGNED-PAYLOAD-TRAILER
  • X-Amz-Meta-Snowball-Auto-Extract: true
  • An Authorization header that includes the access key but may contain a fabricated or invalid signature.
  • The request body contains a tar archive to be auto-extracted.
  • Why exploitation is possible:
  • The handler invokes isPutActionAllowed to extract the access key and evaluate IAM policy, but this step does not validate the request signature.
  • In the subsequent switch rAuthType block, there is no case handling for authTypeStreamingUnsignedTrailer, causing execution to fall through without any cryptographic verification.
  • As a result, any request with a recognized access key and appropriate IAM WRITE permissions is accepted even with an invalid signature, enabling unauthorized data writes.

Why this happens

The underlying issue is an incomplete authentication and authorization flow for a newly introduced authentication mode.

When support for streaming unsigned trailers (authTypeStreamingUnsignedTrailer) was added, it was wired into the standard PutObjectHandler and PutObjectPartHandler but never integrated into PutObjectExtractHandler, which implements the Snowball auto-extract path.

The handler relies on isPutActionAllowed for access control, which extracts the access key and checks IAM policies but does not confirm the authenticity of the request through signature verification.

In the switch rAuthType statement, lack of a dedicated branch for authTypeStreamingUnsignedTrailer means the code path bypasses any signature validation altogether, exemplifying CWE-287: Improper Authentication and CWE-306: Missing Authentication for Critical Function in the context of a sensitive object-write operation.

Potential impact

  • Attackers can upload arbitrary tar archives and have their contents auto-extracted as objects into targeted buckets.
  • Data integrity of buckets can be compromised through unauthorized overwrites, defacement, or insertion of malicious payloads.
  • Stored application assets, backups, or artifacts can be poisoned (e.g., for supply-chain compromise or data corruption).
  • Logs, models, or datasets stored in MinIO may be silently altered, impacting analytics, AI workloads, and downstream decision-making.
  • In multi-tenant environments, one tenant's access key could be abused to affect shared buckets where they have WRITE permissions.

General mitigation guidance

Effective mitigation focuses on restoring robust authentication for this path and limiting the blast radius of any misuse:

  • Enforce signature verification for all auth types:
  • Ensure that every S3-compatible PUT handler, including specialized paths like Snowball auto-extract, explicitly validates cryptographic signatures for all recognized X-Amz-Content-Sha256 modes.
  • Avoid fall-through behavior in authentication-type switch statements; make missing cases fail closed rather than succeed without checks.
  • Harden access key management and IAM policies:
  • Avoid well-known default credentials; rotate access keys regularly and treat them as sensitive secrets.
  • Restrict s3:PutObject permissions to tightly scoped and trusted principals, ideally limited per bucket or prefix.
  • Apply least privilege to service accounts and applications, granting only the minimal WRITE access they require.
  • Reduce exposure of risky request modes:
  • Consider disallowing or strictly controlling STREAMING-UNSIGNED-PAYLOAD-TRAILER usage, especially for endpoints that implement special behaviors such as auto-extraction.
  • Prefer the signed variant (STREAMING-AWS4-HMAC-SHA256-PAYLOAD-TRAILER) in client configurations and document it as the default choice.
  • Network and gateway controls:
  • Use reverse proxies, API gateways, or WAFs to enforce additional request-level policies, such as rejecting certain headers or methods on sensitive paths.
  • Segment MinIO administration and data planes from untrusted networks; limit who can directly reach S3 endpoints.
  • Defense in depth on data integrity:
  • Enable versioning, object locking, or replication policies that make unauthorized modifications easier to detect and recover from.
  • Employ integrity checks or signing on critical objects (e.g., application artifacts, models) so consumers can verify authenticity independently.

How teams detect issues like this

Detection can combine configuration review, traffic inspection, and behavioral monitoring:

  • Configuration and code review:
  • Examine handler implementations for all S3 auth modes to confirm that cryptographic signature checks are consistently applied.
  • Flag any logic where IAM policy checks are performed without subsequent or prior signature validation, especially in switch/if chains keyed on authentication type.
  • Use static analysis rules or linters to identify critical operations (like PUT uploads, auto-extract, or import functions) that lack explicit authentication and integrity verification steps.
  • Runtime traffic analysis:
  • Monitor access logs for requests with X-Amz-Content-Sha256: STREAMING-UNSIGNED-PAYLOAD-TRAILER, particularly when combined with X-Amz-Meta-Snowball-Auto-Extract: true.
  • Alert on anomalous write patterns such as unexpected tar-based uploads, unusual bucket targets for a given access key, or high-volume object creation from unfamiliar clients.
  • Policy and gateway enforcement:
  • Implement WAF or API gateway rules that detect and optionally block specific high-risk headers or content types associated with streaming unsigned payloads.
  • Integrate automated policy checks into deployment pipelines to ensure that insecure authentication modes are not enabled unintentionally.
  • Integrity and anomaly monitoring:
  • Track unexpected changes to critical buckets using checksum comparisons, object versioning, or integrity-monitoring tools.
  • Set alerts on WRITE operations performed by accounts not normally associated with bulk ingest or Snowball-style import workflows.

Sources

SourceLink
github.comgithub.com/minio/minio/commit/76913a9fd5c6e5c2dbd4e8c7faf56ed9e9e24091
github.comgithub.com/minio/minio/pull/16484
github.comgithub.com/minio/minio/security/advisories/GHSA-9c4q-hq6p-c237

Frequently asked

questions

Any attacker who knows a valid MinIO access key with WRITE permissions can exploit this issue over the network, without needing the corresponding secret key or a valid signature, as long as they can reach the MinIO endpoint.

No. The attacker only needs an access key associated with any identity that has s3:PutObject (or equivalent WRITE) permissions on the target bucket; administrative rights are not required, though broader permissions increase impact.

Ensure that every code path handling object uploads validates the cryptographic signature for all supported auth types and fails closed when an auth mode is unrecognized or unsupported. In configuration, avoid enabling or relying on unsigned streaming modes for critical operations and use least-privilege IAM policies.

Treat authentication and authorization as separate, mandatory steps for all critical actions: first verify the request's authenticity with a strong signature, then apply fine-grained authorization checks. Use defense-in-depth with static analysis, code review checklists for new auth modes, and gateway policies to block risky request patterns (such as unsigned streaming uploads) by default.

Find out what's already
exploitable in your codebase.