Industry letter says open-weight models are a security advantage for AI defenders

AravSrinivas · x · 2026-07-25

A repost of George Kurtz’s open letter arguing that AI’s future should include both frontier closed models and frontier open-weight models.

The letter says open models are a security advantage because more teams can inspect, test, and harden them.
It frames open weights as important for defenders, who need access to the same capabilities attackers can use.
The attached letter argues that open weights help with competition, control, and AI safety, and warns policymakers not to confuse legitimate model development techniques like distillation with misuse.
The signatories include a wide set of AI and tech companies and organizations, reinforcing that this is an industry-wide policy push rather than a single-company statement.

More from Companies & People