Sony Neural Network Libraries Injection Flaws

Code injection in Sony nnabla through 1.39.0 can execute attacker-supplied Python when a user runs the 'nnabla_cli encode_param' command on a malicious text-format parameter file. The affected releases are no longer supported, no vendor fix is available, and the disclosure references publicly available exploit code; exploitation depends on the operator choosing to process the crafted file, which keeps the assessed risk low.

Code injection in the NNP Converter utility shipped with Sony Neural Network Libraries (NNabla) through version 1.39.0 allows an attacker who supplies a crafted model file to have Python code executed through eval() when a victim converts that file. The unsafe sink is func_set_nnabla_version_decorate in python/src/nnabla/utils/converter/utils.py, which passes the model's attacker-controlled 'function.type' value into eval(); the CVSS 4.0 vector requires user interaction (UI:P), so nothing happens until a person or an automated pipeline actively loads and converts the malicious model. The CNA reports that exploit code has been released publicly and that the vendor did not respond to disclosure; the library is end-of-life with no fixed version, and the overall priority is 'monitor' because exploitation depends on someone processing untrusted input rather than on a remotely reachable service.