TI-2026-36017 CVE-2026-48746 · GHSA-94f4-hr76-p5j6
vLLM: OpenAI auth bypass
Priority HIGH 78
vLLM is an inference and serving engine for large language models (LLMs).
RHSA-2026:69469
Red Hat AI Inference 3.4.5 (rocm)
Released 2026-09-21 · Red Hat rating Important · View on Red Hat →
Fixed in
- Red Hat AI Inference Server 3.4 registry.redhat.io/rhaii/vllm-rocm-rhel9
Fixes 20 records
TI-2026-42349 CVE-2026-59204 · GHSA-vjc4-5qp5-m44j
Pillow JPEG2000 tiled decode retains a growing scratch buffer and can be used for denial of service
Priority HIGH 77
Pillow is a Python imaging library.
TI-2026-04577 CVE-2026-5241 · GHSA-fgcw-684q-jj6r
Policy Bypass in LightGlue Nested Config Resolution in huggingface/transformers
Priority HIGH 76
A vulnerability in the LightGlue model loading path of huggingface/transformers version 5.2.0 allows an attacker-controlled model repository to execute arbitrary code during model initialization.
TI-2026-42344 CVE-2026-59197 · GHSA-xj96-63gp-2gmr
Pillow: Heap out-of-bounds write in Pillow `ImageFilter.RankFilter` via integer overflow in `ImagingExpand`
Priority HIGH 75
Pillow is a Python imaging library.
TI-2026-42350 CVE-2026-59205 · GHSA-9hw9-ch79-4vh6
Pillow: Controlled heap out-of-bounds write in `ImageCmsTransform.apply()` via output mode mismatch
Priority HIGH 72
Pillow is a Python imaging library.
TI-2026-42347 CVE-2026-59200 · GHSA-jjj6-mw9f-p565
Pillow: Decompression Bomb DoS via PdfParser.PdfStream.decode()
Priority HIGH 72
Pillow is a Python imaging library.
TI-2026-42346 CVE-2026-59199 · GHSA-6r8x-57c9-28j4
Pillow: Heap out-of-bounds write `Image.paste()` / `Image.crop()` via signed coordinate overflow
Priority HIGH 72
Pillow is a Python imaging library.
TI-2026-39958 CVE-2026-55380 · GHSA-phj9-mv4w-65pm
Pillow GdImageFile decompression bomb protection bypass
Priority HIGH 72
Pillow is a Python imaging library.
TI-2026-39957 CVE-2026-55379 · GHSA-45hq-cxwh-f6vc
Pillow BdfFontFile`: `Image.new()` called without `_decompression_bomb_check()` — bomb…
Priority HIGH 72
Pillow is a Python imaging library.
TI-2026-39137 CVE-2026-54234 · GHSA-8wr5-jm2h-8r4f
vLLM: Remote DoS in vLLM via Invalid Recovered Token Reinjection
Priority HIGH 72
vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs.
TI-2026-38994 CVE-2026-54060 · GHSA-5x94-69rx-g8h2
Pillow: `FontFile.compile()`: `Image.new()` called without `_decompression_bomb_check()`
Priority HIGH 72
Pillow is a Python imaging library.
TI-2026-04816 CVE-2026-5497 · GHSA-wcwg-c5fc-9vrc
Unbounded Frame Count in video/jpeg Base64 Data URL Processing Leads to OOM DoS in vllm-project/vllm
Priority HIGH 72
vLLM versions 0.8.0 and later are vulnerable to an Out-of-Memory (OOM) Denial of Service (DoS) attack due to unbounded frame count processing in the `VideoMediaIO.load_base64()` method.
TI-2026-38992 CVE-2026-54058 · GHSA-62p4-gmf7-7g93
Pillow: Out-of-bounds read via attacker-controlled row stride on Pillow's mmap path (McIdas AREA files)
Priority HIGH 69
Pillow is a Python imaging library.
TI-2026-35886 CVE-2026-48526 · GHSA-xgmm-8j9v-c9wx
PyJWT: Public-key JWK accepted as HMAC secret enables forged HS256 tokens when mixed families are allowed
Priority HIGH 68
PyJWT is a JSON Web Token implementation in Python.
TI-2026-51646 CVE-2026-73556 · GHSA-48jh-3gj7-fg8v
Regular expression denial of service in vllm-project vllm
Priority MEDIUM 63
vLLM is an inference and serving engine for large language models.
TI-2026-32202 CVE-2026-44223 · GHSA-83vm-p52w-f9pw
vLLM: extract_hidden_states speculative decoding crashes server on any request with penalty parameters
Priority MEDIUM 62
vLLM is an inference and serving engine for large language models (LLMs).
TI-2026-39183 CVE-2026-54283 · GHSA-82w8-qh3p-5jfq
Starlette: request.form() limits silently ignored for application/x-www-form-urlencoded enable DoS
Priority LOW 39
Starlette is a lightweight ASGI framework/toolkit.
TI-2026-40113 CVE-2026-55574 · GHSA-rwxx-mrjm-wc2m
vLLM: ReDoS via structured_outputs.regex compiled without timeout in xgrammar and outlines backends
Priority LOW 38
vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs.
TI-2026-30106 CVE-2026-41523 · GHSA-q8gq-377p-jq3r
vLLM: Security Check Bypass via assert Statement in Activation Function Loading Allows Arbitrary Code Execution
Priority LOW 37
vLLM is an inference and serving engine for large language models (LLMs).
TI-2026-27214 CVE-2026-34993 · GHSA-jg22-mg44-37j8
AIOHTTP Vulnerable to Deserialization of Untrusted Data
Priority LOW 31
AIOHTTP is an asynchronous HTTP client/server framework for asyncio and Python.