Depth First Labs joins Open Secure AI Alliance to preserve open-weight model access for security

andreamichi · x · 2026-08-11

Depth First Labs announced it is a founding member of the Open Secure AI Alliance, alongside NVIDIA and other industry leaders, working to preserve access to open-weight models for security.

The alliance argues that restricting access to open-weight models would make defenders dependent on a small number of closed providers, reducing competition and increasing pricing power, making frontier-level security too expensive to deploy broadly. Open-weight models enable security companies to train specialized models and provide economical frontier-level security. Depth First Labs demonstrated this potential with dfs-large1, the first model not trained by a frontier lab to achieve best-in-class performance on vulnerability detection tasks.

Original post →

More from Safety

Safety channel →