Disrupting malicious uses of AI | February 2026By OpenAI News / February 25, 2026 Our latest threat report examines how malicious actors combine AI models with websites and social platforms—and what it means for detection and defense.