These last few days we have observed in Extremadura the umpteenth chapter starring the fraudulent use of AI. With an application, a group of minors (there are up to 10 involved) created the fictitious nude of young women, also minors and flesh and blood, to then spread it. The action represents an attack on the honor, image and privacy of the girls (in addition to a violation of the law on the protection of minors). The question is: is it possible to avoid another case like this?
Those who understand the matter consulted by La Vanguardia move between moderate optimism and slight skepticism. Albert Agustinoy, Cuatrecasas partner and director of the firm’s intellectual property and technology group, understands that the future approval of the European regulation on the regulation of AI may represent some progress, albeit with nuances. He explains that one of the novelties that have been incorporated in the latest revision is about generative AI (the one that generates content), a fact that he identifies as positive. He assures that there is even reference to deepfakes (generated video, image or audio that imitates the appearance and sound of a person). But it also detects a flaw: it doesn’t explicitly make a pornographic video generation tool that includes third-party images illegal.
“The answer he gives is not absolutely clear, to my surprise”, he admits. The rule – continues – establishes the obligation of the software developer and of the hosting platforms (web hosting) that the technology created detects the combination, for example, of the image of a minor (also of an adult) with the request for pornographic content, “something that will obviously end up infringing someone’s rights”. “It is not so much a matter of banning this kind of tools, which could also run into the right to freedom of expression, but of preventing them from generating content that will undoubtedly be completely illegal under EU law “, add.
He understands – “without knowing the final version of the regulation”, he emphasizes – that these kinds of tools will be obliged, in some way, to take a series of safeguards to avoid cases similar to the one in Extremadura, a fact that he values ??positively .
Carlota Planas, a lawyer specializing in intellectual property and new technologies, also values ??it in a certain way, for whom “these tools should be approved before they see the light of day”, a fact that does not currently happen “because there is no regulations in force”.
Consequently, he sees with good eyes the future approval of the European joint, but he understands that the developers will explore all possible corners. In wich way? “They will add to the prospectus of the program (as if it were a drug) undesirable side effects for which they do not want to be held responsible”.
In any case, he defends that the institutions are a little late. “AI is run amok. It is a very easy tool to use, and more so for digital natives. In addition, it is global and exponential: it only takes one click, which can be done by millions of people in the same instant”.
It also highlights that this is not just any tool: it has enormous potential and the possibility of not leaving much of a trace. “You can be on the other side of the world and create an AI application without anyone noticing,” he warns.
He remembers, however, that the fraudulent use of AI “can be prosecuted administratively, civilly and even criminally”. Its dissemination can also constitute a crime. Regarding this, adds Agustinoy, the emergence of AI has not meant any legal novelty. “You only have to remember the famous case of the actor Santi Millán”, he argues.
It is true, however, that the perpetrators of the Extremadura case, most of them under 14 years of age, are unimpeachable. “They are not criminally responsible even if they have committed a crime. There may be an administrative infraction, resulting in a fine, which parents should be responsible for,” concludes Agustinoy.