{"id":"CVE-2026-93989","published":"2026-09-19T23:17:10.043","lastModified":"2026-09-19T23:17:10.043","description":"vLLM through 0.29.0 fails to properly validate bad_words token indices against the model's generation output width in SamplingParams.update_from_tokenizer(). Attackers can supply out-of-bounds token indices that corrupt logits memory of concurrent requests, causing different in-flight HTTP requests to return incorrect tokens.","cvssScore":3.1,"cvssSeverity":"LOW","cvssVector":"CVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:U/C:N/I:L/A:N","cwes":["CWE-129"],"vendors":[],"products":[],"references":[{"url":"https://github.com/vllm-project/vllm","tags":[]},{"url":"https://github.com/vllm-project/vllm/blob/98dff2a81d747d1dba01a47f939f48c3526d4206/vllm/sampling_params.py#L694-L753","tags":[]},{"url":"https://github.com/vllm-project/vllm/blob/98dff2a81d747d1dba01a47f939f48c3526d4206/vllm/v1/worker/gpu/sample/bad_words.py","tags":[]},{"url":"https://github.com/vllm-project/vllm/pull/48824","tags":[]},{"url":"https://www.vulncheck.com/advisories/vllm-through-0.29.0-cross-request-logits-corruption-via-bad-words","tags":[]}],"exploitRefs":[{"url":"https://github.com/vllm-project/vllm","tags":[]},{"url":"https://github.com/vllm-project/vllm/blob/98dff2a81d747d1dba01a47f939f48c3526d4206/vllm/sampling_params.py#L694-L753","tags":[]},{"url":"https://github.com/vllm-project/vllm/blob/98dff2a81d747d1dba01a47f939f48c3526d4206/vllm/v1/worker/gpu/sample/bad_words.py","tags":[]},{"url":"https://github.com/vllm-project/vllm/pull/48824","tags":[]}],"hasPoc":true,"ai":null}