On June 15, 2023, The European Parliament, a main legislative branch of the European Union, adopted the draft law known as the AI Act by an overwhelming majority. The world’s first law on AI ensures full compliance with the EU laws and values, including security, privacy, transparency, non-discrimination, as well as social and environmental well-being. However, there are potential risks connected to the regulation, such as AI startups moving their operations to unregulated countries.

What is AI Act about

The European Union’s AI Act is supposed to be the world’s first regulatory framework for building AI products. Although the draft proposal is still in the works and is heavily negotiated, the framework outlines several risk categories. According to them, different AI products require different regulations:

  • Most prominently, the draft introduces bans on biometric surveillance, emotion recognition, and predictive policing AI systems.
  • AI systems that can harm people’s health, security, rights or influence voters in political campaigns and in recommender systems used by social media platforms are also considered high-risk. They would have to register in the EU database and comply with severe limitations.
  • Generative purpose AI would have to guarantee robust protection of fundamental rights and law compliance. Such models like ChatGPT would also be required to provide higher transparency, revealing that content was generated by AI and providing safeguards against illegal content generation.
  • In order to boost AI innovation, the law adds exemptions for research activities and AI components provided under open source licenses. It promotes the so-called regulatory sandboxes to test artificial intelligence before its implementation.
  • The AI Act also enforces the role of the European Artificial Intelligence Authority, who will be tasked with monitoring how the AI ​​rulebook is implemented.

Why is it important?

  • In general, the law is designed to restrict big tech companies from harming people’s privacy and security. However, it’ll still apply to Europe’s smaller startups and can cause their accelerated growth to stall, reducing the innovative potential of the European region.

“Startups will go to the US, they’ll develop in the US, and then they’ll come back to Europe as developed companies, unicorns, that’ll be able to afford lawyers and lobbyists. Our European companies won’t blossom, because no one will have enough money to hire enough lawyers,”

Piotr Mieczkowski, managing director of Digital Poland, commented to Sifted.
  • Those companies that don’t comply with the law will risk facing fines of up to €30 million, or 6% of the company’s total worldwide annual turnover. So, evidently, there is still uncertainty whether the law will be beneficial or hinder the AI development in Europe.