Deepfake services advertised on dark web at low cost, Kaspersky warns
Kaspersky's research team discovered advertisements on the dark web offering deepfake creation services at low cost—starting at $50 for fake videos and $30 for voice messages. The services claim to provide real-time synchronisation of facial expressions with speech and voice cloning capabilities, though many are suspected to be scams.
What happened
Kaspersky's Global Research and Analysis Team (GReAT) discovered advertisements on the dark web promoting deepfake services for video and audio with real-time synchronisation between face, voice and text. The advertised services start at $50 for fake videos and $30 for fake voice messages, with prices increasing based on content type and duration.
The advertisements offer multiple deepfake options including facial feature modification during video calls via messaging or meeting applications, facial modification during verification processes, and live camera feed replacement with synthetic content. The promoters claim to provide software that synchronises facial expressions with spoken text even in foreign languages, plus voice cloning and voice tone and pitch modification tools to reflect different emotions. Many of these advertisements are suspected to be fraudulent schemes targeting potential buyers.
Kaspersky noted that these low-cost services represent a significant decrease from previously tracked deepfake services on the dark web, which cost between $300 and $20,000 per minute. Some platforms now offer malicious language models designed from scratch to run locally on devices without reliance on public models.
Reported impact
Kaspersky expressed concern that threat actors are actively experimenting with AI tools and integrating these technologies into their operations, potentially enabling extortion, piracy, and impersonation of officials, public figures and ordinary individuals.
Index analysis
Kaspersky recommended cybersecurity experts double their efforts to counter these challenges, with AI-enhanced security expert capabilities and defence effectiveness as effective countermeasures. The firm advised organisations to ensure effective cybersecurity systems with qualified IT experts, use threat intelligence platforms to detect potential threats, conduct regular employee awareness training on deepfake dangers, and teach staff to identify deepfake indicators such as jerky movements, lighting variance between frames, unnatural skin tone, unusual or absent eye blinks, image distortions, and deliberately low-quality video or poor lighting conditions.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.