AI Deepfakes Intensify Scam Threat as Most Americans Report Exposure

Generative AI has lowered the cost of producing convincing fake video, cloned voices and personalized phishing messages, allowing criminals to impersonate relatives, companies and public agencies at scale. Deepfakes are eroding trust in digital communications and weakening familiar defenses based on spotting awkward language or visual flaws, raising the risk of identity theft, account takeovers and unauthorized financial transactions.

A Pew Research Center report released July 31, 2025, found that 73% of U.S. adults had experienced at least one type of online scam or attack. The survey showed 48% had encountered fraudulent credit- or debit-card charges, 29% had a personal account hacked and 24% had surrendered information after a scam message. Separately, 68% said wider AI use would make online scams and attacks more common.