Lawmakers in the European Parliament have approved the AI Act, which aims to regulate AI systems based on a risk-based approach. The law passed with an overwhelming majority of 523 votes in favour, 46 against, and 49 abstentions. This marks a historic day as it is the first regulation in the world that provides a clear path for safe and human-centric development of AI. The act will most likely enter into force in May and will be overseen by national authorities, supported by the AI office within the European Commission.
Machine learning systems will be categorized into four main groups based on the potential risk they pose to society. Systems that are considered high risk will be subject to stringent rules before they can enter the EU market. The general-purpose AI rules will come into effect one year after the entry into force in May 2025, while the obligations for high-risk systems will be implemented in three years. Member states will be responsible for setting up national oversight agencies, as they have 12 months to nominate these watchdogs.
In response to the approval of the AI Act, Cecilia Bonefeld-Dahl of Digital Europe emphasized the need to do more to support companies based in Europe. Currently, only 3% of the world’s AI unicorns come from the EU, with significantly more private investment in AI in the US and China. It is crucial for European companies to tap into the growing global AI market, which is expected to reach $1.5 trillion by 2030, without facing unnecessary obstacles.
Ursula Pachl of the European Consumer Organisation (BEUC) welcomed the law’s approval, noting that it will help consumers join collective redress claims if they have been harmed by an AI system. While the legislation could have been more protective of consumers, the focus now shifts to the European Commission and national governments demonstrating their commitment to implementing and enforcing the AI Act promptly. This will require providing relevant regulators with the necessary resources to ensure compliance.
The EU is looking to work with partner countries to ensure a global impact of the AI Act rules. There is an openness to collaborating with other nations to promote and implement these regulations while building a governance system with like-minded parties. The implementation and compliance of businesses and institutions with the AI Act will be key moving forward. Additionally, lawmakers are also working on further legislation for the next mandate, including potentially a directive on conditions in the workplace and AI. The AI office within the European Commission will provide oversight and support to ensure the effective implementation of the rules.
