On May 19, the European Commission released draft guidelines on determining whether an artificial intelligence (AI) system qualifies as high-risk. under the provisions of the EU AI Act, as reported by Hunton Andrews Kurth LLP. As outlined within Article 6 of the Act, the proposed guidelines categorize AI systems based upon potential impact to physical well-being, safety, or fundamental rights.
Although the proposed draft guidelines establish a general framework for AI providers to evaluate and classify their respective AI systems, specific information related to the criteria used to identify such classifications has been left out in publicly available sources. The European Commission created the proposed guidelines to provide a uniform basis for the interpretation and enforcement of Article 6 of the Act among all member states of the European Union.
Timing Is Important for Video AI Developers in Europe Who Must Comply with the Requirements of the EU AI Act
As the phased roll-out of the EU AI Act continues, identifying those AI systems classified as high-risk will be critical for determining which video AI developers are required to comply with certain requirements including conducting risk assessments, obtaining certifications from third parties, etc.
Get the latest model rankings, product launches, and evaluation insights delivered to your inbox.
Industry analysts have begun to examine which types of AI systems may require additional operational effort resulting from the implementation of the new regulatory framework. Friction is expected due to new regulatory compliance requirements relating to AI safety, ethics and regulatory compliance.
AI developers who are subject to regulations that designate them as high-risk will need to comply with significant documentation requirements, perform conformity assessments, and undergo continuing monitoring activities prior to deploying their systems into markets. AI systems that are designated as high-risk are also subject to stringent requirements for accuracy, reliability, and transparency prior to deployment.
In this case, it appears that the European Commission is still accepting input from interested stakeholders. Final versions of these proposed guidelines will likely influence future regulatory actions taken against AI platforms within each of the EU’s 27 member states. Other countries are observing how Europe develops its regulatory approach towards AI in anticipation of developing similar regulatory frameworks themselves.