AI news story
Cisco Foundation AI Releases Antares: 350M and 1B Open-Weight Models That Localize Known Vulnerabilities Inside Real Codebases
Cisco Foundation AI has released Antares, a family of small language models trained to pinpoint where known vulnerabilities l…
Editor's take
Cisco Foundation AI has introduced Antares, a pair of open-weight language models specifically engineered to identify the precise locations of known security flaws within software code.
This development is significant because it offers a more accessible and potentially more accurate tooling for developers and security teams aiming to proactively address vulnerabilities. By achieving a File F1 score of 0.209 on the Vulnerability Localization Benchmark, Antares-1B outperforms previous benchmarks like GLM-5.2, suggesting an advancement in the practical application of LLMs for code security.
Future developments to monitor include the real-world adoption rate of Antares by major software development organizations and its performance against zero-day vulnerabilities. The effectiveness of these models in integrating seamlessly with existing CI/CD pipelines will also be a key indicator of their long-term impact on secure software development practices.