Regulatory

Global: EU AI Act Needs Revision to Ensure Genuine Risk-Based Approach, Report Suggests

0
EU AI Act Needs Revision to Ensure Genuine Risk-Based Approach, Report Suggests
Share this article

A recent legal study has added to the growing criticism of the European Union’s Artificial Intelligence Act, highlighting that crucial provisions do not adhere to a truly risk-based approach. This world-first attempt at comprehensive AI regulation could lead to “legal uncertainty, potential over-regulation, and unjustified increases in compliance costs,” according to the report. However, it also identifies opportunities for improvement.

Martin Ebers, president of the Robotics & AI Law Society (RAILS) in Germany and a professor of IT Law at the University of Tartu, Estonia, acknowledges that the AI Act includes tools to support future-proof legislation that can be implemented in line with a genuine risk-based approach. Ebers emphasizes the need for the Act to be applied as originally intended, with adjustments to ensure clarity and effectiveness.

The AI Act, which received final approval in May, categorizes AI applications into four risk levels, with stricter rules applied to higher-risk uses. This framework aims to balance the protection of rights with the encouragement of innovation. However, Ebers, who is also a senior research fellow at Humboldt University of Berlin, suggests that the Act should incorporate sector-specific legislation to avoid inconsistencies and over-regulation.

Ebers recommends that the European Commission AI Office should “clarify, concretize, amend, supplement, modify, or even delete a large number of provisions” to keep pace with rapid technological advancements and changing AI applications.

Since the draft was first published in 2021, the AI Act has been the subject of intense debate. Human rights organizations have criticized it for insufficient protections, while business groups have expressed concerns about heavy compliance burdens. The latest report, commissioned by the Computer & Communications Industry Association (CCIA Europe), underscores these issues.

Boniface de Champris, CCIA Europe’s senior policy manager, warns that key provisions of the AI Act fail to follow a genuine risk-based approach. He advises legislators worldwide to be cautious about adopting the EU’s framework without addressing its structural shortcomings to avoid replicating these issues globally.

Share this article

Global: Mastercard to Phase Out Manual Card Entry for Online Payments in Europe by 2030

Previous article

Global: Apple Delays AI Features in EU Due to Digital Markets Act Concerns

Next article

You may also like

Comments

Comments are closed.

More in Regulatory