Members of the European Parliament have voted for new laws that are much stricter towards ChatGPT and other AI models.
The European Parliament recently debated regulating the use of artificial intelligence. In a new regulation called the “AI Act”, members of Parliament established several future rules concerning the dangers and remedies to be adopted to regulate the monumental rise of artificial intelligence like Midjourney or ChatGPT.
The world’s most famous AI model, ChatGPT, is obviously at the forefront of these new bans. It will now be required that OpenAI’s little robot respects European rules on copyright and the publication of abstracts.
The vote was also eagerly awaited concerning “deepfakes”, these images or videos generated by artificial intelligence and which are often used for disrespectful or completely degrading montages. This content must now be labeled as such under penalty of punishment or fine.
The AI Act provides for several levels of risk depending on the artificial intelligence you use in the future. AI that threatens citizens’ rights is now prohibited, including the untargeted retrieval of facial images from the Internet or CCTV footage to create facial recognition databases. Any type of profiling of a person will also be prohibited as well as solutions that manipulate human behavior or exploit people’s vulnerabilities.
The first law on artificial intelligence was passed with a large majority of favorable opinions. Debate for several years, its very first version proposed in 2021 (before the explosion of the ChatGPT phenomenon) had extremely strict frameworks. A version that did not please certain heads of state like Emmanuel Macron who declared that such a text would make Europe an uninteresting market for artificial intelligence players. The risk is that specialized companies will turn more towards the United States to invest in technology.