Response errors via predictable prefix-cache hashes
Published Mar 4, 2025 · Updated Mar 4, 2025
Insufficient randomness in AIBrix 0.2.0 allows attackers to disrupt subsequent responses via crafted prompt hash collisions. The prefix cache hashes prompt token blocks with xxhash64 under a fixed seed, making outputs predictable and allowing colliding inputs to reuse or pollute cached prefixes. Exploitation requires the ability to submit chosen prompts and construct a collision; the reported consequence is erroneous subsequent responses, not code execution or data disclosure.
Summary
What happened
Insufficient randomness in AIBrix 0.2.0 allows attackers to disrupt subsequent responses via crafted prompt hash collisions. The prefix cache hashes prompt token blocks with xxhash64 under a fixed seed, making outputs predictable and allowing colliding inputs to reuse or pollute cached prefixes. Exploitation requires the ability to submit chosen prompts and construct a collision; the reported consequence is erroneous subsequent responses, not code execution or data disclosure.
The record
- CVE
- CVE-2025-1953
- Published
- Mar 4, 2025
- Updated
- Mar 4, 2025
- Vendor
- vLLM Project
- Product
- AIBrix
- Classifications
- CWE-330, T1565.001
- Attack vector
- network
- Privileges
- Unavailable
Timeline
How it unfolded
- Mar 4, 2025CVE publishedPublication date reported by the CVE source.
- Mar 4, 2025Record updatedLatest update available in the CVE record.
Exploitability
Present is not the same as exploitable
Compare your product and version with the public record. A matching version still requires validation against your environment.
Is a vulnerable build present?
Compare these published version ranges with your installed build and any vendor patches.
- Affected versionversion=0.2.0
What conditions does exploitation require?
What is affected?
Published CVSS scores
No CVSS assessment is available in this record.
CVSS describes severity. EPSS estimates exploitation probability.
Attacks
What attackers are doing with it
Daily unique IPs observed by Shadowserver honeypots for known exploited vulnerabilities (KEVs). Missing observations do not establish an absence of attacks.
Weakness, pattern, technique
Public exploit references
No public exploit references are available in this record.
Labels summarize the accepted research assessment. They do not indicate a test against your environment.
Technologies
Your stack
See the directory against your own environment.
Your stack
Check the software in your environment
Book a demo to see how Hinoki identifies affected software and validates exploitability in your environment.
Book a demo