Disrupting Malicious AI Uses: How OpenAI Is Fighting Back in 2026 From AI-powered romance scams to state-backed influence campaigns, OpenAI’s latest report exposes how malicious AI operations work and why most of them quietly fail.