Our Director of New Technologies and AI, Juan Pablo González, spoke with Diario Financiero about the bill that seeks to protect individuals’ digital identity and the challenges posed by its implementation.
Last week, the Chamber of Deputies approved in general terms—in the first constitutional reading—the bill regulating the creation and dissemination of realistic digital imitations of people’s images, bodies, or voices generated using artificial intelligence (AI), known as deepfakes.
It was approved by a vote of 128 in favor, two against, and five abstentions. Since amendments were introduced, the bill will return to the Committee on the Future, Science, Technology, Knowledge, and Innovation for a detailed vote.
The motion was introduced in the Lower House in August 2025 by a group of deputies and former deputies from the Frente Amplio (FA) led by Gael Yeomans, and in April of this year, the Committee on the Future forwarded the report with the final approved text to the Chamber.
The bill consists of 12 permanent articles and one transitional article. Broadly speaking, it seeks to protect individuals’ digital identity and integrity against content created or altered using AI that simulates their image, body, or voice, and applies to both individuals and legal entities—and, in particular, to “digital social media platforms that offer their services within the national territory.”
Yeomans, a member of the Future Commission, noted that during the legislative process at this stage, the concept of “plausible but false digital representations” was incorporated into the text, which allows for “specifically penalizing content that is misleading and damages one’s reputation,” he said.
The proposal stipulates that individuals or entities that violate the law could face legal action ranging from being held liable for damages caused to the adoption of precautionary measures—including the removal of content at the request of the affected party—and other measures aimed at “preventing further harm when there is an imminent danger of irreparable harm.”
It also grants enforcement powers to the future Personal Data Protection Agency, which will be able to impose fines ranging from 5,000 UTM ($358 million) to 10,000 UTM ($716 million) for the benefit of the state on both legal entities and individuals.
In addition, it imposes specific obligations on digital social media platforms, such as designating a legal representative domiciled in Chile, implementing a reporting channel, and using “visible labels” for content generated or modified using AI.
The bill excludes content “that is clearly identifiable as satire, parody, criticism, or artistic expression regarding the person depicted digitally,” that does not cause deception, confusion, or harm to one’s reputation, and is “clearly identified as content generated and/or altered using AI or other technology.”
Yeomans noted that the bill is “urgent,” and that one aspect that could be regulated with “greater precision” during the detailed debate is the use of AI by the government and public institutions.
“Cases such as the alteration of images using AI to exaggerate the progress of the trench along the northern border show that it may be necessary to establish stricter standards,” the congresswoman stated.
Penalties
Congressman Andrés Celis (RN), a member of the Future Commission, described the initiative as “necessary” and said that “it establishes a general principle that individuals have the right to digital integrity.”
However, he noted areas for “improvement” in subsequent legislative stages, such as “better defining certain aspects of civil liability and the system of penalties, adjusting them to the magnitude of the violation,” he said.
He also noted that a distinction must be made between the scope of application for legal entities and individuals, and that it is important to avoid both “excessive” compliance costs and any inhibition of the “legitimate use and development of these technologies,” as well as to prevent regulatory overlap in light of a future AI framework law.
Meanwhile, Juan Pablo González, Director of New Technologies and AI at az, welcomed the fact that the bill establishes regulations on deepfakes, as there are currently no specific regulations on this topic in Chile.
However, he noted that the bill focuses on the use of AI-generated content but does not address risk management or the various actors involved in the process, such as developers or providers.
Digital Platforms
Regarding the obligations for social media platforms, González noted that requiring a legal representative in Chile could be “complex” for platforms without a local presence, while reporting channels could become a “tremendously costly” mechanism without clear rules governing their operation.
He commented that the bill does not distinguish between violations of varying severity nor does it incorporate mitigating or aggravating factors. Nor does it specify whether the future Personal Data Protection Agency, in addition to having the authority to impose sanctions, will also oversee digital platforms, nor does it provide for specific administrative procedures.
González noted that the bill also includes “criteria for the Agency to determine the amount of the fine on a case-by-case basis,” which “could be counterproductive in certain scenarios” because it might end up imposing similar penalties for violations of different nature.