The use of artificial intelligence to clone voices and create fake videos has raised concerns about the protection of a person’s identity. In response, Italian Prime Minister Giorgia Meloni has requested to register her voice as a sound mark with the European Union Intellectual Property Office, sparking a debate on the limits of legal tools in protecting a person’s identity from artificially created content.
In Colombia, violations of the personal data protection regime can result in administrative sanctions of up to 2,000 minimum monthly legal salaries, equivalent to $3,502 million, as well as the suspension of activities related to data processing and, in cases provided by law, the closure of operations.
Legal Framework
The Law 2502 of 2025 established that the penalty for personal falsity can increase by up to one-third when artificial intelligence is used, as long as the conduct does not constitute another crime. José Luis Jérez, a lawyer specializing in data and image, explained that a person’s voice and face can be protected by different provisions, depending on how they are obtained, processed, and used.
Read Also: Military Situation in Cauca Worsens Amid Drone and Gun Attacks
Jérez noted that the Law 1581 of 2012, which establishes the general regime for the protection of personal data, is relevant when the information allows a person to be identified or made identifiable and is subject to operations such as collection, storage, use, circulation, or suppression. This means that the analysis is not limited to determining whether a voice or image was copied, but also whether personal information was obtained without meeting legal requirements.
Deepfakes and Falsity
The Law 2502 of 2025 modified the article 296 of the Penal Code, related to personal falsity, and defines deepfakes as false audiovisual records, including photographs, videos, images, or sound recordings, created with artificial intelligence to appear authentic regarding a person’s speech or behavior.
Francisco Bernate, a criminal lawyer, stated that the reproduction of a voice or face using artificial intelligence does not constitute a crime in itself. The penal relevance depends on the use made of the material and whether the conduct meets the elements of an offense provided by law.
Read Also: Colombian courts expose flaws in fiscal liability rules
They noted that an imitation created for entertainment purposes, for example, cannot be automatically equated with content designed to impersonate another person and obtain money or harm them. Bernate explained that the situation may be different when a video or audio is used to request transfers, deceive family members or colleagues, impersonate a company executive, or support a fraudulent maneuver.
Penalties and Sanctions
The penalties and fines will depend on the conduct accredited and the applicable norms; there is no single sanction that covers all scenarios of misuse of artificial intelligence. However, if there are indications of impersonation or another crime, the affected person can go to the Fiscalía General de la Nación to report the facts to the authorities.
When the problem involves the processing of personal data, it is also possible to examine the possibility of filing a complaint with the SIC, according to the competencies of that entity. While there are tools available, the solution requires identifying how the content was created, with what purpose it was used, and what rights were compromised.
