This period saw high-urgency, election-focused misuse of synthetic media and growing operational fraud risks. U.S. authorities warned that adversaries are actively using convincing deepfake audio and video against election and campaign officials, alongside a separate FBI alert about AI-generated impersonation schemes targeting voters and donors. View Advisory View Advisory A new study reported that major AI models are now more willing to create fake election imagery, while Spain’s opposition deployed an AI-generated political deepfake video during an active election period. View Article View Article

Financial and identity-verification ecosystems remain under pressure. Reporting shows underground sellers offering full KYC bypass kits with synthetic documents and deepfake videos to defeat live verification. View Article Voice-clone fraud continued to produce real-world harm, including a Florida law-enforcement impersonation scam and a family emergency clone-voice case in Indiana, while INTERPOL launched a public campaign highlighting escalating deepfake scam risks. View Article View Article View Advisory

Detection and platform defenses show mixed progress. New research indicates baseline deepfake detectors degrade sharply on newer generators and can be driven below 2% accuracy under attack, Supporting Report while provenance signals remain inconsistent across AI images. View Study On the defensive side, Microsoft announced Teams integrations for third‑party deepfake detection and Google rolled out Android protections against spoofed calls and messaging scams. View Article View Article

State-linked influence operations persisted, with OpenAI reporting disruptions of Russian and Iranian false-front entities using AI-generated personas to seed content to real outlets. View Report Overall, the threat environment is assessed as worsening given concurrent escalation across elections, financial fraud, identity verification bypass, and demonstrated degradation of legacy detection.