CVE-2026-15605
A security vulnerability has been detected in wandb 0.25.2.dev1. Affected is the function ArtifactManifestEntry.download in the library wand
CVSS
3.1
Low
EPSS
0.2%
p13
KEV
—
Exploit Today
4
0-100
Published: Jul 13, 2026 · Last modified: Jul 14, 2026 · CWE-327 · CWE-328
0.2%EPSS · 30 days0.2%
2026-09-222026-10-19
A security vulnerability has been detected in wandb 0.25.2.dev1. Affected is the function ArtifactManifestEntry.download in the library wandb/sdk/lib/hashutil.py of the component Artifact Integrity Validation. The manipulation leads to use of weak hash. The attack may be initiated remotely. A high degree of complexity is needed for the attack. The exploitability is told to be difficult. The pull request to fix this issue awaits acceptance.
- github.comhttps://github.com/wandb/wandb/
- github.comhttps://github.com/wandb/wandb/issues/12030
- github.comhttps://github.com/wandb/wandb/pull/12031
- vuldb.comhttps://vuldb.com/cve/CVE-2026-15605
- vuldb.comhttps://vuldb.com/submit/855675
- vuldb.comhttps://vuldb.com/vuln/378114
- vuldb.comhttps://vuldb.com/vuln/378114/cti
CVECVSSEPSSKEVRExploitTitleMod.
CVE-2026-31615.9 MED—
——0IBM Cloud Pak System 2.3.6.0, and 2.3.6.1 uses weaker than expected cryptographic algorithms that could allow an attacker to decrypt highly sensitive information.1dCVE-2026-69303.7 LOW—
——0IBM Concert 1.0 through 3.0.0 uses weaker than expected cryptographic algorithms that could allow an attacker to decrypt highly sensitive information.1dCVE-2026-591166.8 MED28.8%
——9Use of a broken or risky cryptographic algorithm in Windows Remote Desktop allows an unauthorized attacker to bypass a security feature over a network.7dCVE-2026-651047.1 HIG3.0%
——1NVIDIA Megatron-LM for all platforms contains a vulnerability in inference, where an attacker could cause hash collisions in the cache used for inference requests. A successful exploit of this vulnerability might lead to data tampering and partial information disclosure.7dCVE-2026-974689.1 CRI7.8%
——2Apache CXF's STSTokenValidator and Security Token Service (STS) cached validated security tokens under a non-cryptographic 32-bit hash of the token (Java Arrays.hashCode/hashCode()), and treated a cache hit as proof that the presented token had already been validated. An attacker could craft a token (for example a UsernameToken or a self-signed SAML Assertion) whose hash collides with a cached entry. The token would then be accepted without password validation, signature trust verification or a call to the STS. This could let the attacker authenticate as another user and, through STS token validation or renewal, obtain STS-signed tokens for that identity.
Users are recommended to upgrade to versions 4.2.4 or 4.1.9 or 3.6.13, which fix this issue.12hCVE-2026-1028244.3 MED3.1%
——1Russh is a Rust SSH client and server library. Prior to 0.63.0, the hybrid ML-KEM 768 and X25519 implementation in russh/src/kex/hybrid_mlkem.rs accepts an all-zero 32-byte peer X25519 public key in both server_dh and compute_shared_secret, forcing the X25519 contribution to the combined shared secret to zero. A malicious SSH peer can therefore make the combined secret depend only on ML-KEM, defeating the hybrid exchange's intended fallback protection if ML-KEM is later weakened. This issue is fixed in version 0.63.0.21d