The Spanish government gives the green light to the law that will fine if it is not indicated if a photo is made with AI

MadridOne day after Pope Leo XIV put artificial intelligence (AI) in the spotlight of his first encyclical, the most relevant papal document, the Spanish government has sent the organic law for the good use and governance of AI to Congress. The Council of Ministers this Tuesday gave the green light in a second reading to the text that adapts Spanish legislation to the European AI regulation approved on March 13, which seeks to ensure an ethical, inclusive, and beneficial use of artificial intelligence and to enable users to clearly differentiate which content is real and which is virtual.

In fact, one of the keys to the regulation is that once it comes into force (the parliamentary process is just beginning now), companies behind synthetic content, that is, content created with AI, that do not label it as such may be fined. It should be recalled that from next August 2, it will be mandatory to identify audiovisual content created with AI, as stipulated in the European regulation, but no sanction will be received in case of non-compliance. This will happen once the member state's regulation, for example the Spanish one, comes into force.

Spanish Prime Minister Pedro Sánchez, who precisely this Tuesday, and amid the political storm over the Plus Ultra case, is traveling to Rome to meet with the Pope before his visit to Spain, has echoed Leo XIV's words. "The encyclical calls us all to account [...]. AI is not neutral, and digital power can lead us to new atrocities if it is not oriented towards the common good," Sánchez said through a message on X. Sánchez has always championed the need for a global response to the "challenge" of AI, especially because the direction is decided by a "small and reduced group of companies," as he has denounced on repeated occasions.

The Minister for Digital Transformation and Public Function, Óscar López, also defended the Pope's vision on AI at a press conference this Tuesday because it "protects" human rights, and contrasted it with the vision of the great "techno-oligarchs".

Extraterritorial law

The rule is extraterritorial. This means that it affects all AI providers that offer a service in Spain, regardless of where they offer it from. For example, an AI with a North American seal that operates from the United States will have to comply with this regulation when it operates in the Spanish state.

The core elements of the rule are: the transparency of the algorithms used, the responsibility of the executives behind AI companies, and the protection of minors, as detailed by the minister.

Prohibited AI

In the same way as in the community regulation, Spanish law establishes prohibited practices for artificial intelligence, such as the use of subliminal techniques to manipulate decisions without consent. Also, the use of chatbots that are capable of identifying possible gambling addicts to encourage them to enter online gaming platforms or exploit vulnerabilities related to age, disability, or the socioeconomic situation of individuals, among others.

As a novelty, deep fakes (hyperrealistic audio or video content) of a sexual nature and the generation or alteration of child pornography are prohibited. Non-sexual deep fakes will remain legal, but they will have to be clearly labeled with the acronym AI (the English acronym for artificial intelligence), as mentioned previously.

Sanctioning regime

The law provides for different levels of offenses, and sanctions are set that range from 6,000 euros to 35 million euros in the most serious cases. For example, the marketing of some of the artificial intelligence systems that are prohibited. Public administration will continue to be exempt from fines, but sources from the ministry assure that the formulas for admonishment will be effective. This means, for example, that if the police use the real-time biometric identification system, which the law also prohibits, they will not be fined.

Control

The Spanish Data Protection Agency will continue to be responsible for the management and processing of biometric data, while AI affecting justice will be the responsibility of the General Council of the Judiciary (CGPJ). Regarding other AI use cases, the entity responsible for controlling them will be Aesia, an organization that will have about fifty analysts responsible for monitoring the correct application of artificial intelligence in the State in accordance with the regulations.

The final text approved this Tuesday arrives one year after its approval in the first round. The draft law has been submitted to public hearing, a period during which allegations could be submitted, and has been reviewed by various bodies, including the General Council of the Judiciary.