← Back

Wire · founder news, decoded · technology

Cisco open-sources AI models for bug hunting

Published

21 July 2026

Topic

technology

Sectors

AI InfrastructureCybersecurityDeveloper Tools

Geography

United States

Source

Read at axios.com

Verified

Fusion42 · 21 July 2026 · Fusion42 review

Cisco open-sources two small AI models (Antares-350M and Antares-1B) designed to scan codebases for vulnerabilities, running 500 repositories in 15 minutes for under $1 versus $100+ for larger models like GPT-5.5. The models can run locally, keeping sensitive source code on-premise, and are released on Hugging Face with vetting restrictions to prevent adversary access.

This Wire brief sits within Fusion42's coverage of AI Infrastructure, Cybersecurity and Developer Tools. Wire is Fusion42's founder-focused intelligence feed: each story is connected to the funds and startups it names — every one with a live profile on Raise or Scout — so founders can follow the capital and the momentum behind the headline rather than just the headline itself. Wire analysis is one of the live surfaces Arthur reasons over.

The Wire takeaway

If you're selling vulnerability scanning or code security tools, you now compete with free open-source models that cost $1 to run instead of $100, and your customers can deploy them internally without sending code anywhere. Your moat just became narrow unless you're faster, more specialised, or already embedded in their pipeline.

Related on Wire

Topics

AI Infrastructure · Cybersecurity · Developer Tools · open-source-models · vulnerability-scanning · cost-efficiency · on-premise-ai · security-tooling