The European Parliament has adopted a majority position on the regulation of artificial intelligence, paving the way for the world’s first comprehensive AI law. The AI Act, as it’s known, has taken a significant step towards final approval, marking a milestone in the global tech industry.
On June 14, 2023, the European Parliament voted in favor of AI regulation with an overwhelming majority of 499 votes for, 28 against, and 93 abstentions. The AI Act, a comprehensive project aimed at regulating AI, has made a fundamental stride towards its final approval.
The AI Act: A Potential Global Model
The AI Act aims to protect fundamental civil rights and safeguard health and safety from potential AI threats while simultaneously fuelling technological innovation. The AI Act is being hailed as a potential model for policymakers worldwide, given its pioneering approach to AI regulation.
The AI Act adopts a “risk-based” approach, banning AI applications that pose an unacceptable risk and imposing a stringent regime for high-risk use cases. The new law also addresses the need for transparency with the material used to train AI operating in Europe, indicating whether the material used is copyrighted.
The Act has significant implications for generative AI models, like OpenAI’s GPT-4, ChatGPT etc. These foundational models will face new transparency requirements. The Act also mandates that companies disclose the copyrighted data used to train these models.
The Act imposes an obligation for mandatory labeling of the content generated by AI. It also requires companies to have a model to prevent users from generating illegal content, a move that has sparked debate about user control and search monitoring.
The AI Act primarily focuses on categorizing AI operating in Europe by risk, following an evaluation. It categorizes certain AI as posing an unacceptable risk and thus will be prohibited from use in Europe. These include AI that manipulates people’s behavior, classifies people based on behavior, social status, or personal characteristics, and systems for remote real-time biometric identification.
Following the Parliament’s vote, a final version of the law will be reviewed among representatives of the three branches of the European Union — the European Parliament, the European Commission, and the Council of the European Union. A definitive agreement is expected by the end of the year.
The AI Act is a testament to Europe’s commitment to balancing technological innovation with the protection of fundamental rights. As the world watches, the Act could set a precedent for global AI regulation, shaping the future of AI and its role in society.