Election-related synthetic media threats intensified, with U.S. federal authorities issuing back-to-back public advisories warning of convincing deepfake audio/video and AI‑enabled phishing targeting campaigns and election officials View Advisory and active fraud schemes impersonating campaigns and officials to solicit donations and personal data View Advisory. A separate report indicates a pending U.S. FCC action could loosen political robocall restrictions, potentially normalizing AI voices in last‑minute electoral communications Source. Spain also saw a high‑profile deepfake‑style political video from a major opposition party ahead of national voting View Article.

Criminal misuse continued to drive large financial and public‑safety impacts: reports describe a deepfake video‑call executive impersonation leading to roughly $25M in enterprise losses in Hong Kong View Article, an estimated €95M voice‑clone fraud affecting a major bank in Italy View Article, and cloned‑voice vishing that impersonated a Florida sheriff to extort residents View Article.

On the national‑security front, OpenAI reported disrupting covert “false‑front” influence operations that used AI personas and fabricated materials to infiltrate media and online platforms Source. Meanwhile, law‑enforcement action in France reportedly targeted deepfake CSAM and non‑consensual image sites, indicating ongoing large‑scale synthetic exploitation networks.

Defensive and provenance developments advanced, including new documentation for verifying C2PA‑signed image credentials Source, Android device‑level AI defenses against scam calls and messages View Article, and newsroom authenticity workflows combining C2PA and robust watermark recovery View Article. The U.S. judiciary was also warned about deepfake evidence risks, highlighting rising legal‑system exposure View Article.