llama.cpp b9592 updates LibreSSL for secure local AI inference

The popular open-source project llama.cpp, known for enabling local LLM inference on consumer hardware, has released version b9592. This release focuses on security by updating the bundled LibreSSL library from an earlier version to 4.3.2. The commit is signed with GitHub's verified signature (GPG key ID B5690EEEBB952194) and authored by Adrien Gallouët from Hugging Face. The update applies to all platform builds—macOS (Apple Silicon ARM64 and Intel x64, with optional KleidiAI acceleration), Linux (x64, ARM64, s390x, plus Vulkan, ROCm 7.2, OpenVINO, and SYCL FP32), Windows (x64 CPU and ARM64, plus CUDA 12/13, Vulkan, and HIP), Android ARM64, and even openEuler builds.

While the change itself is incremental, it underscores the project's commitment to maintaining robust security in the AI inference stack. With over 116,000 stars on GitHub and 19,500 forks, llama.cpp is a cornerstone of the local AI movement. Users can download prebuilt binaries or compile from source directly from the release page. For developers running sensitive workloads or embedding llama.cpp in applications, this LibreSSL patch is a recommended update. The project continues to be actively maintained, with frequent releases enhancing stability and security.