Election-related deepfake activity remains the week’s most consequential risk. The FBI and CISA issued a public advisory warning that adversaries are actively using synthetic audio and video to target U.S. election and campaign stakeholders, heightening risks of social engineering, compromise, and mis/disinformation during the 2026 cycle. Complementary reporting highlights AI-powered voter deception including fake fundraising, misleading guidance, and amplification of a Russian-linked deepfake narrative in the midterms.

Criminal abuse continues across voice-cloning fraud, KYC evasion, and large-scale scams. Researchers documented an underground market selling AI-generated videos and real-time face manipulation to defeat KYC checks, while phishing-as-a-service offerings are now integrating voice AI agents that increase social-engineering potency.

OpenAI also reported dismantling a Cambodia-based scam operation that used AI for fake personas, translations, and forged materials.