The European Parliament's Internal Market Committee and Civil Liberties Committee have adopted a negotiating mandate to create the first artificial intelligence (AI) legislation in the EU.
The goal is to ensure that AI systems are “safe, transparent, traceable, inclusive and environmentally friendly.” Restrictive measures have been proposed to limit the use of AI in facial recognition or social credit systems. In addition, the definition of an AI system has been worked on to accommodate not only current systems, but also those of the future.
There are many risks associated with the use of high-risk artificial intelligence for the European Union. The proposed European Parliament and Council Regulation on harmonized rules for AI raises a number of questions. Who is affected? Providers, users and suppliers employing or providing an AI system in the EU. Any AI system employed by companies under EU jurisdiction, which includes U.S. systems employed directly or indirectly, will also be impacted by this new legislation.
AI systems will be classified according to their level of inherent risk. Harmful uses of AI will be prohibited as they contravene EU values, as will systems that negatively impact the safety of individuals or their fundamental rights. Low-risk AI systems, such as email spam filters, will be exempt from this legislation. On the other hand, high-risk systems, such as biometric-based systems used by law enforcement, or migration and border control systems, will require a compliance assessment and be subject to a set of obligations.
Generative artificial intelligence will also be subject to this legislation. Although not systematically considered a high-risk system, it will have to ensure transparency that the content is generated by an AI system and not by humans. Finally, the legislation will take into account environmental protection, so that AI systems are also environmentally friendly.
This new legislation will be put to a vote in June 2023, before the final terms are agreed between representatives of the EU Parliament, Council and Commission. Those affected will then have two years to comply with the regulation. With this new legislation, the EU will be able to closely supervise the use of AI and ensure that systems are safe, transparent, traceable, inclusive, and environmentally friendly.
I'm a big fan of short stories about people – I'm a pro at tech and smartphones, serial literature, and writing in my spare time.