OpenAI - Disrupting Malicious Uses of Our Models

This is the February, 2026 report from OpenAI that discusses their work in disrupting malicious use of their models...