Election security dominates today’s risk picture. U.S. federal advisories warn that threat actors are actively using deepfake audio and video alongside AI-enhanced phishing and smishing to target campaigns and election officials, and that AI-generated impersonation is being weaponized for election-themed fraud against voters and donors. Security Advisory Security Advisory

Coordinated influence and disinformation operations persist across platforms. OpenAI reports disrupting Russian- and Iranian-linked false-front operations using fabricated documents, personas, and scripted propaganda workflows; AP separately highlights a China-linked “Spamouflage” network posing as U.S. voters on X. Supporting Report View Article

Criminal abuse continues beyond politics: UK police warn of more than 1,300 AI-enabled investment endorsement scams, while Florida authorities report repeated cloned-voice impersonation of a sheriff’s captain in arrest-warrant schemes. View Article View Article

On defense, provenance and platform safeguards are advancing: OpenAI and Microsoft document C2PA/watermark-based provenance checks, Reuters and Sony demonstrate near-live C2PA workflows, and Google announces Android defenses against voice-cloning and messaging scams. Vendor Advisory Vendor Advisory View Article View Article

Net assessment: With multiple High-urgency election incidents, ongoing state-aligned influence operations, voice-clone frauds, and limited time for countermeasures before voting peaks, the threat environment is worsening.