The year 2025 witnessed a massive leap in deepfake technologies, with the accuracy of simulating faces, voices, and movements reaching levels that make it difficult to distinguish between reality and fake.
Estimates indicate that the number of fake videos will rise from about 500,000 in 2023 to about 8 million in 2025, with an even greater increase expected in 2026.
Developments include:
* Video models that maintain consistency of movement and identity, allowing different movements to be combined with multiple identities.
* Sophisticated voice cloning from just a few seconds, allowing it to be used in fraud.
* Consumer tools such as “Sora 2” and “Veo 3” have made the production of sophisticated fake content accessible to everyone.
With the increasing quantity and accuracy of fake content, its detection has become more difficult, leading to harassment, misinformation, and fraud. Technology in 2026 is expected to move towards the production of instant videos that accurately simulate human behavior, with source authentication and infrastructure protection serving as the primary line of defense against forgery.
