“Generated by AI”, the mandatory mention in Europe from this Tuesday

The reform is the first visible change derived from the new European Artificial Intelligence Regulation, or IA Act, as it is already known in the European media. With this new standard, Europe aspires to become a global benchmark for AI regulation, given that its investments in the development of this technology, compared to those of the United States and China, both private and public, are infinitely lower.

The European Commission announced last week the creation of a fund of 30,000 million euros to build six large data centers on the continent when the United States for the same thing is approaching an investment that is close to a trillion euros.

The new transparency obligations do not prohibit using artificial intelligence to write texts, create images or videos or respond to customers with virtual assistants, but they do force the citizen to know that they are seeing something created by a machine or interacting with a virtual assistant and not a human.

The new regulation has implications for companies that own artificial intelligence tools. It forces them to incorporate technical mechanisms, such as metadata or invisible signatures, whichthat allow us to identify the origin of the content that is created on their platforms.

The regulation focuses mainly on so-called ‘deepfakes’, videos, images or audio manipulated to make it appear that a person said or did something that they never said or did. From now on they must be clearly identified to avoid deception and to stop misinformation and political manipulation.

Not all texts will escape the norm. Those who report on matters of public interest and have not been supervised by humans should make this clear. This section generated criticism because some media accused the European Commission of try to encourage censorship. Brussels defends itself by saying that it does not censor anything, that it only tries to identify materials that were written with an AI and never went through human control, so that citizens are informed and take the necessary precautions.

In addition to the media, the norm affectsto communication and advertising agencies, marketing departments, digital platforms of any size or public administrations. Any organization capable of producing hundreds or thousands of documents, images or automated responses each day.

The regulation is accompanied by significant fines. In some cases the fines can reach 15 million euros or 3% of global income. of the offending company. It is the same sanctioning model that the European Union approved when it launched the General Data Protection Regulation.

Since it was launched, notices about cookies on web pages have become widespread. It was then, as now, according to the European Commission, increase transparency on the internet without slowing down innovation or censorship.

Critics say that Brussels has given birth to a mouse and that in a few weeks citizens will be so used to seeing warnings that content has been generated by AI that they will begin to not pay attention to it. The European Commission defends itself by saying that at least citizens They will be able to know if content is human or artificial creation.