The Rise of Deepfakes
Deepfake technology has emerged as a powerful but controversial tool. A deepfake is an audiovisual manipulation technique that uses artificial intelligence and deep learning to create highly realistic synthetic content. Through sophisticated algorithms, deepfakes can superimpose faces, alter expressions, clone voices and generate body movements, making it possible to…

Deepfake technology has emerged as a powerful but controversial tool. A deepfake or “deep falsification” is an audiovisual manipulation technique that uses artificial intelligence and deep learning to create highly realistic synthetic content.
Using sophisticated algorithms, deepfakes can superimpose faces, modify expressions, clone voices and generate body movements, allowing the creation of fake videos, images and audio that are hard to tell apart from the real thing.
With the right quick fine-tuning, anyone can create seemingly real images or make the voices of prominent political or economic figures and artists say whatever they want. While creating a deep falsification is not a criminal offense in itself, many governments are moving toward stronger regulation of the use of artificial intelligence to prevent harm to the parties involved.
While more than 400 experts in AI, technology and digital ethics signed an open letter demanding urgent measures against deepfakes, Sumsub's Identity Fraud Report reveals the real impact of this technology across different industries and regions of the world.
The signatories of the letter titled «Disrupting the Deepfake Supply Chain», state that penalties are needed for producing child pornography with this technology, penalties for those who create or spread harmful and false content, and a requirement that software developers and distributors take steps to prevent their products from generating deepfake content.
A survey of 199 cybersecurity leaders who attended the World Economic Forum's Annual Meeting on Cybersecurity in 2023 showed that 46 percent of respondents were most concerned about the *»advance of adversarial capabilities: phishing, malware development, deepfakes»* in terms of the risks that artificial intelligence poses to cybersecurity in the future.
According to SumSub's State of Deepfakes 2023 report, the proliferation of fake and misleading videos produced with AI systems increased by 550% between 2019 and 2023. Sumsub Research detected a tenfold increase in the number of deepfakes worldwide across all industries from 2022 to 2023.
For example, fraud incidents in the Philippines rose by as much as 4,500% year over year, followed by countries such as the United States, Vietnam and Belgium. These figures are alarming and reflect how quickly this technology is developing and spreading.
One of the most worrying aspects of the rise of deepfakes is their impact on the pornography industry. Lewd, fictitious content created with AI accounts for 98% of all deepfake videos online, an increase of 400% over the past year. In addition, 99% of those targeted were women, which points to a recurring pattern of technology-enabled gender-based violence.
Deepfakes also pose a significant threat to the identity of individuals and companies. The firm Onfido estimates that deepfake fraud increased by 3,000% last year. This trend is especially worrying ahead of electoral processes, since 51% of the world's population will take part in elections in the near future. Deepfakes have the potential to manipulate public opinion and undermine the integrity of democratic processes.
Despite the experts' call, strategies to regulate the development and use of AI face significant challenges. Current laws do not adequately define or limit the production and distribution of deepfakes, and the requirements imposed on creators are often ineffective. In addition, regulatory initiatives are developing in isolation, which undermines efforts to address this problem.
Experts propose the widespread adoption of tamper-proof cryptographic digital seals on photographs and videos as a practical solution.
These seals would make it possible to verify the authenticity of content using open-source applications. It is also suggested that device manufacturers, software developers and media companies work together to popularize these content authentication methods.
As deepfakes continue to advance, the market valuation of AI is expected to exceed 300 billion dollars in 2025, which underscores the importance of addressing the challenges associated with deepfakes.
Governments, companies and society at large must work together to develop effective regulatory frameworks and technological solutions that protect the integrity of information and identity in an increasingly digitalized world.
The rise of deepfakes poses unprecedented challenges for society in the digital age. While experts advocate urgent measures to address this threat, the reality revealed by the Identity Fraud Report shows the scale of the problem.
Only through joint efforts and innovative solutions can we navigate the complex landscape of deepfakes and safeguard the core values of our society.