Deepfake quality and volume surge dramatically in 2025
AI-generated synthetic media improved dramatically in 2025, with deepfake volume increasing from roughly 500,000 in 2023 to approximately 8 million by 2025. Voice cloning and video generation models now produce content indistinguishable from authentic recordings, and consumer tools have made deepfake creation accessible to the general public, with reports of large-scale fraud including over 1,000 AI-generated scam calls per day to retailers.
What happened
Over the course of 2025, deepfakes improved dramatically in quality and volume. AI-generated faces, voices and full-body performances that mimic real people increased in quality far beyond expert expectations, and were increasingly used to deceive people.
The volume of deepfakes grew explosively: DeepStrike estimates an increase from roughly 500,000 online deepfakes in 2023 to about 8 million in 2025, with annual growth nearing 900%. Video generation models now maintain temporal consistency with coherent motion and stable faces without the flicker or distortions that once revealed deepfakes. Voice cloning has crossed what researchers call the 'indistinguishable threshold,' requiring only a few seconds of audio to generate a convincing clone. Consumer tools including OpenAI's Sora 2, Google's Veo 3, ChatGPT and Gemini have pushed technical barriers almost to zero, allowing anyone to generate polished audio-visual media in minutes.
Reported impact
For many everyday scenarios—especially low-resolution video calls and media shared on social media platforms—deepfakes are now realistic enough to reliably fool nonexpert viewers. In practical terms, synthetic media have become indistinguishable from authentic recordings for ordinary people and, in some cases, even for institutions.
Voice cloning capability is already fueling large-scale fraud, with some major retailers reporting over 1,000 AI-generated scam calls per day.
Index analysis
Technical shifts underlying the escalation include video generation models designed to maintain temporal consistency, producing videos with coherent motion and consistent identities while eliminating perceptual tells that once gave away deepfakes. Voice cloning has eliminated the perceptual tells that once gave away synthetic voices. Consumer-accessible tools have democratized the ability to create deepfakes, with AI agents capable of automating the entire process from script generation to final output.
Classification
Sources and evidence
Sources not yet published
This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.