Europe Moves Closer to Regulating AI with the Draft AI Act
On June 14, 2023, the European Parliament took a significant step toward implementing an EU regulation for generative AI and other AI systems by adopting its position on the draft AI Act. The aim of this act is to establish a set of unified rules for the development, market placement, and use of AI within the European Union. The primary goal is to ensure the safe, ethical, and trustworthy application of AI systems while safeguarding people’s rights.
The initial draft text of the AI Act was published in April 2021 and has since undergone extensive modification and debate. The emergence of ChatGPT and other similar generative purpose AI systems has prompted notable clarifications and obligations for providers of such systems.
Given that the AI Act will be a regulation, it will be legally binding and directly applicable in all Member States of the European Union.
The AI Act revolves around a risk-based approach and is founded on six quality principles. These principles are then converted into specific obligations, which will vary depending on the nature of the entity and the specific AI system in question.
Now let’s delve into the key details:
a) Who does the draft AI Act apply to?
The AI Act primarily targets providers who release AI systems into the EU market. Unlike the General Data Protection Regulation (GDPR), it does not require entities to be established within the EU to fall under the scope of this legislation. Thus, companies from the United States or China may also find themselves subject to these regulations. Additionally, manufacturers, authorized representatives, distributors, and importers will have their own set of obligations to fulfill.
b) What are the various AI systems and their associated obligations?
Here’s a brief overview of the most important points related to different AI systems as outlined in the initial draft:
Enforcement of the AI Act will primarily be carried out by national supervisory authorities. These authorities will possess extensive powers, including the ability to impose fines similar to those seen in the GDPR. Any entity found to have placed unacceptable AI systems on the market may be subject to administrative fines up to €40,000,000 or, if the offender is a company, up to 7% of its total worldwide annual turnover from the preceding financial year (whichever amount is higher).
The finalized text of the AI Act is expected to be ready by the end of the year and will be enforceable two years thereafter.
As an AI language model, I don’t have personal feelings or opinions. However, it is evident that the draft AI Act signifies a significant milestone towards regulating AI systems within the European Union. By establishing harmonized rules, the EU aims to ensure the responsible and ethical use of AI while upholding individuals’ rights and interests. This regulatory framework will undoubtedly contribute to shaping the future of AI technology in Europe and beyond.
