Trustworthy Artificial Intelligence (AI) systems possess the potential to bestow countless advantages, but they also harbor a multitude of risks. A clear regulatory framework becomes indispensible.
May 2023, the European Parliament voted the Commission’s proposal to establish new regulations ensuring that AI systems used in the EU are safe, transparent, ethical, unbiased and under human control. This marks a significant milestone, paving the way for the commencement of final negotiations, known as the trilogue, among the Commission, Parliament, and the Counsel to refine the legislation’s ultimate form.
A risk-based approach
Starting from the point of view of product safety, the AI Act adheres to a risk-based approach whereby specific rules shall be applied contingent upon the classification.
Prohibited practices
The list of Prohibited practices comprises all those AI systems whose use is considered unacceptable as contravening the European Union values. These include government-operated social scoring mechanisms and voice-assisted toys that promote hazardous conduct among children.
High and Limited Risk AI systems
The crux of the AI Act centers around the regulation of the High Risk category, encompassing domains such as recruitment and medical devices. These activities are permitted, however subject to adherence to AI requirements and pre-implementation conformity assessments. On the other hand, AI systems categorized under Limited Risk, such as bots and deep fakes, will be permitted with the caveat of -amongst others- information and transparency obligations.
This Act will furnish us with horizontal high-level requirements that will necessitate the supplementary formulation of business standards. Will your company demonstrate compliance with the fundamentals of the AI Act, thereby imbuing your product with the trustworthiness for success in the market? We are here to guide you to compliance.
