Europe Proposes Guardrails on AI

• Proposed regulations governing AI in Europe will have worldwide implications when adopted.
• Organizations should begin locating, classifying, and documenting applications that could be targeted by the regulations.
   

 

In Apr. 2021, the European Commission released a draft of regulations intended to control the use of AI technologies in the European Union (EU). The Artificial Intelligence Act is designed on a risk-based framework, with increasing amounts of controls placed on uses of AI deemed to pose higher risk of physical harm or human rights. Like the General Data Protection Regulations (GDPR), the AI Act is likely to form the basis of similar regulations outside of the EU. Organizations using technologies targeted by the act should understand its implications now, although many open issues remain, particularly with respect to assessing the level of risk and quality of training data sets.

Goals of the Artificial Intelligence Act

The stated goals of the Artificial Intelligence Act (formally, “Proposal for a Regulation of the European Parliament and of the Council Laying Down Harmonised Rules on Artificial Intelligence”) are twofold: first, to encourage innovation in AI within the EU and, at the same time, to place “guardrails” on AI-enabled applications.