And this commit contained a refactoring from BROTLI_BOOL to BrotliDecoderErrorCode so the situation was that probably one of the error conditions previously simply returned false but needed more detail about what went wrong to continue safely.
Well either that or this change from `if (prev_ix < cur_ix)` to `if (prev_ix < cur_ix && cached_backward <= max_backward)`.
Worth checking how brotli actually reaches your stack. In Node it's bundled into the runtime itself (zlib's brotli support), so an audit of your package tree won't surface it the way a normal dependency would - the fix arrives with a runtime release, not a package bump. The Python package on PyPI is already at 1.2.1. Until your runtime picks up the patched brotli, refusing br content-encoding at the edge seems like the fastest mitigation.
And this commit contained a refactoring from BROTLI_BOOL to BrotliDecoderErrorCode so the situation was that probably one of the error conditions previously simply returned false but needed more detail about what went wrong to continue safely.
Well either that or this change from `if (prev_ix < cur_ix)` to `if (prev_ix < cur_ix && cached_backward <= max_backward)`.
https://x.com/AikidoSecurity/status/2113759110625906950
The python package has already been updated to 1.2.1. https://pypi.org/project/brotli/1.2.1/
https://nitter.jaydenha.uk/OpenAI/status/2113766354386084338
and this is the actual CVE report, which doesn't say much: https://www.cve.org/CVERecord?id=CVE-2026-111497
whether a hello wolrd in fastapi is affected might also be interesting
That's pretty difficult if you make outbound requests, and don't already have a TLS interception proxy set up.