EU Parliament Approves Artificial Intelligence Act

The European Parliament approved today the Artificial Intelligence act, the world’s first AI legislation aiming to put safeguards on a new type of technology that’s been evolving at a very fast pace in recent years. The new regulation, which has been discussed since 2021 and still needs to be fully adopted by the EU Parliament, aims to encourage the development of AI technology that respects fundamental rights.

“Artificial intelligence is already very much part of our daily lives. Now, it will be part of our legislation too,” said Roberta Metsola, President of the European Parliament today.

In practice, the AI Act will introduce some obligations for AI technologies based on their potential risks. For starters, the General-purpose AI (GPAI) systems that enabled the creation of apps like ChatGPT will need to meet certain transparency requirements and comply with EU copyright laws. There will also be stricter obligations for high-risk AI systems used in critical infrastructure, education, employment, justice, and democratic processes.

The AI Act will also ban AI-powered biometric categorization systems that could threaten citizen rights, though there will be some exceptions for the use of biometric identification systems by law enforcement. The AI act will also forbid the use of emotion recognition technology in the workplace and schools, as well as AI technology that “manipulates human behavior or exploits people’s vulnerabilities.”

The press release from the EU Parliament also says that “regulatory sandboxes and real-world testing will have to be established at the national level, and made accessible to SMEs and start-ups, to develop and train innovative AI before its placement on the market.” While the European Union is the first organization to try to regulate artificial intelligence technology, there are also concerns about how the new legislation could restrict innovation in the European continent.

The generative AI landscape is currently dominated by a couple of American companies including OpenAI, Google, and Microsoft. Earlier this month, we saw Mistral AI, a French startup working on open AI models announce a strategic partnership with the Redmond giant. This pretty much sums up how far ahead Big Tech companies are in the AI race today, and the EU Commission actually said that it would be looking into Microsoft’s partnership with Mistral AI.

The EU’s AI Act will become fully applicable 24 months after its publication, which still leaves a lot of time for companies to figure out how to comply with the legislation. And as we’ve seen recently with the EU’s Digital Markets Act, some companies like Apple will do the bare minimum to comply with EU laws.