MEPs must give the first green light on Wednesday June 14 to the European project to regulate artificial intelligence (AI), a key step in the race started by the legislator in the face of the frenetic pace of ChatGPT-type innovations.
The European Union wants to be the first in the world to adopt a comprehensive legal framework to limit the excesses of AI, while ensuring innovation.
Brussels proposed an ambitious draft regulation two years ago, the examination of which has been delayed and further delayed in recent months due to controversy over the dangers of generative AI capable of creating texts or images.
Obtain voluntary commitments
The European Parliament is due to adopt its position at noon, during a vote in plenary session in Strasbourg. The long-awaited green light should allow negotiations to start with Member States on Wednesday evening to finalize the legislation, if possible, before the end of the year.
Commissioner Thierry Breton, who brought the text with his colleague Margrethe Vestager, asked that the process conclude “during the next few months.” But the regulation will not enter into force before 2026, in the best of cases. Believing there was urgency, the two officials announced their intention to obtain voluntary commitments from the companies as soon as possible.
Of great technical complexity, artificial intelligence systems fascinate as much as they worry. While they can save lives by enabling a quantum leap in medical diagnosis, they are also exploited by authoritarian regimes for mass surveillance of citizens.
The general public discovered its immense potential late last year with the launch of California-based OpenAI’s editorial content generator ChatGPT, which can write original essays, poems or translations in seconds.
“We need rules”
But the spread on social networks of false images, more real than life, created from applications such as Midjourney, has alerted to the risks of manipulation of opinion and the dangers for democracy.
Scientists have called for a moratorium on development of the more powerful systems, until they are better regulated by law.
Parliament’s position broadly confirms the Commission’s approach. The text builds on existing rules on product safety and will impose mainly company-based controls.
The heart of the project consists of a list of rules imposed only on applications considered “high risk”. These would be systems used in sensitive areas such as critical infrastructure, education, human resources, law enforcement, or migration management.
inform users
Among the obligations: provide for human control over the machine, the establishment of technical documentation, or even the establishment of a risk management system. Compliance with them will be supervised by the control authorities of each member country.
The European Parliament intends to take better account of generative AIs of the ChatGPT type by requesting a specific regime of obligations essentially repeating those provided for high-risk systems.
The Commission proposal, presented in April 2021, already provides a framework for AI systems that interact with humans. It will force them to inform the user that they are in contact with a machine, and it will force imaging applications to specify that they were artificially created. An obligation that will probably extend to the texts.
Bans will be rare. They will refer to applications contrary to European values, such as citizen rating systems or mass surveillance used in China.
MEPs also want to remove exemptions provided by the Commission to allow remote facial recognition of people in public places by law enforcement. This topic should fuel discussions with Member States that invoke the fight against crime and terrorism to reject the ban on this controversial technology.
Source: BFM TV
