Deepfakes: a pending conversation
Since 2020, I have devoted eleven articles to exploring the deepfake phenomenon. Eleven attempts to warn, educate and raise awareness about a threat that grows exponentially while our collective response remains, for the most part, stalled. This persistence reflects a deep concern that we are dangerously underestimating a technology that threatens the very foundations of…

Since 2020, I have devoted eleven articles to exploring the deepfake phenomenon. Eleven attempts to warn, educate and raise awareness about a threat that is growing exponentially while our collective response remains, for the most part, stalled.
This persistence reflects a deep concern that we are dangerously underestimating a technology that threatens the very foundations of our shared reality.
The silent escalation
The updated data are striking. Between 2019 and 2023, deepfakes increased 550%, while from 2022 to 2023 they multiplied tenfold globally, according to SumSub.
But behind these statistics lies a digital humanitarian crisis. 98% of deepfake videos are non-consensual pornography, with 99% of victims being women. This goes beyond the technology debate: it is a form of gender-based violence amplified by algorithms.
Deepfake fraud shot up 3,000% last year, according to Onfido. We are not talking about a theoretical risk. It is a reality that is already eroding trust in financial institutions, the media and democratic processes.
The paradox of vulnerable leadership
Our country presents a fascinating paradox. We rank seventh among 33 Latin American nations in government readiness for AI, which positions us as regional leaders in technology strategy.
Yet documented cases show our vulnerability. The impersonation of Central Bank governor Héctor Valdez Albizu, fake audio attributed to President Luis Abinader, and the recent manipulation of broadcaster Roberto Cavada in out-of-context videos show that we are already fertile ground for these digital manipulation techniques.
This contradiction exposes an uncomfortable truth. Technological leadership without robust ethical and legal frameworks is empty leadership. We lack specific legislation against malicious deepfakes, operating in a legal limbo that leaves citizens and institutions exposed.
Anatomy of a triple crisis
I have identified three interconnected dimensions that hinder an effective response.
The data crisis shows in the absence of specific statistics on deepfakes in our context, which prevents the design of evidence-based policies. We cannot fight what we do not measure.
The awareness crisis is reflected in widespread public ignorance about deepfakes: what they are, how they work, how to detect them. This deficit makes us easy targets. Digital literacy is not a luxury; it is a necessity for democratic survival.
The action crisis is evident in the excessive reliance on future legislation, such as Bill 02197-2023 on AI, under review since December, which perpetuates inaction. Meanwhile, the existing legal tools —Article 338 of the Penal Code, Ley 53-07 on High-Tech Crimes, Ley 172-13 on Data Protection— remain underused for lack of specialized training.
Viable paths
Deepfake detection is a costly «arms race», with commercial systems running between $100,000 and $500,000 a year. Still, international models point to possible paths. The European mandatory labeling, Singapore's specific legislation with its emphasis on digital literacy, and Estonia's secure digital identity system offer valuable lessons.
Cryptographic digital seals to authenticate content emerge as a practical, preventive solution, potentially more effective than purely reactive strategies.
Let's put a foot in the future
Deepfakes challenge our ability to tell truth from fiction and to preserve the social trust that sustains democracy.
Our response must be multidimensional. Robust digital education, specific legal frameworks, investment in authentication technologies, and multisector collaboration are the pillars of an effective strategy. We must reject toxic innovation and consciously steer it toward the common good.
How much more harm will we tolerate before we act?