BSI Compliance Navigator: latest white paper- Ethical and trustworthy Artificial Intelligence
In 2021 the European Commission released the draft AI Act (AIA) (EC, Artificial Intelligence Act, 2021). Following the initial release, the AIA text has changed and developed as it has made its way through the EU legislative process. In Nov 2022, the Commission adopted its position to take into negotiations with the European Parliament (Council, Nov 2022). The AIA is expected to become a regulation early in 2024, and is expected to come into application in 2027, according to the latest text, where the transition period is defined as 3 years.
Once the AIA becomes law, as a regulation it will have immediate impact throughout the EU. AIA is expected to have a global effect, not because it is the first legal framework impacting AI – there are already other frameworks, standards and guidelines already published - but because it is not restricted to the geographical territory of the EU. The AIA clearly states (Article 2, 1a) that the scope covers “providers placing on the market or putting into service AI systems in the Union, irrespective of whether those providers are established within the Union or in a third country” and “(1c) providers and users of AI systems that are located in a third country, where the output produced by the system is used in the Union”. A similar effect to the GDPR - which has seen widespread global adoption - is expected by the application of AIA.
Read the full white paper, written by Dr. Aris Tzavaras, Head of Artificial Intelligence, Notified Body, BSI Group, to find out more about how the draft AIA interprets
• Definition of AI (Art.3) Prohibited Artificial Intelligence Practices
• Classification (Art. 6)
• Requirements for High-Risk AI
• General Purpose AI Systems (Art. 4a-c)
• Overview of main obligation for High-Risk AI
• Conformity assessment process of High-Risk AI systems (Art.43)
• Role of Standards and Common Specifications (Art. 40-41)
• Role of Notified Bodies
