GitHub Advisoryadvisory · PyPI
GHSA-5qmp-p3c4-72qj MLflow: Deterministic sampling in dataset digest enables predictable collisions
LOW
Official page ↗
- Aliases
- CVE-2026-10803
- Ecosystem
- PyPI
- Packages
- mlflow
- Severity
- Low
- CVSS vector
- CVSS:3.1/AV:L/AC:H/PR:L/UI:N/S:U/C:N/I:L/A:L
- Weaknesses
- CWE-327
- Reviewed
- 2026-07-15
Affected packages
• PyPI mlflow: < 3.10.1 → fixed in 3.10.1
Description
A flaw has been found in MLflow up to 3.10.0. This issue affects the function mlflow.data.digest_utils of the file mlflow/data/digest_utils.py of the component Dataset Digest Computation. This manipulation causes use of weak hash. It is possible to launch the attack on the local host. The attack is considered to have high complexity. The exploitability is assessed as difficult. The exploit has been published and may be used. The project was informed of the problem early through a pull request but has not reacted yet.
Related
Alias · CVE (1)
- CVE-2026-10803 A flaw has been found in MLflow up to 3.10.0. vulnerability
Weakness · CWE (1)
- CWE-327 Use of a Broken or Risky Cryptographic Algorithm weakness
References
- https://nvd.nist.gov/vuln/detail/CVE-2026-10803 advisory
- https://github.com/mlflow/mlflow/issues/22419 web
- https://github.com/mlflow/mlflow/pull/22420 web
- https://github.com/mlflow/mlflow package
- https://github.com/pypa/advisory-database/tree/main/vulns/mlflow/PYSEC-2026-195.yaml web
- https://vuldb.com/cve/CVE-2026-10803 web
- https://vuldb.com/submit/831462 web
- https://vuldb.com/vuln/368252 web
- https://vuldb.com/vuln/368252/cti web