On July 29, 2021, the Centre for Information Policy Leadership (“CIPL”) at Hunton Andrews Kurth submitted its response to the European Commission’s Consultation on the Draft Artificial Intelligence Act (the “Act”). Feedback received as part of this consultation will feed into discussions with the European Parliament and the European Council as the proposal makes its way through the EU legislative process.

The Draft AI Act incorporates several recommendations made in CIPL’s paper on Adopting a Risk-Based Approach to Regulating AI in the EU, including the Act’s approach to regulating only high-risk AI, the definition of AI based on uses and not rigid sectors, the reliance on industry self-assessment to demonstrate product conformity, the establishment of an AI innovation board and the inclusion of a statutory basis for regulatory sandboxes.

However, in CIPL’s view the AI Act does not appear to sufficiently account for imperatives such as:

  • Providing for outcomes-based rules;
  • Enabling organizations to calibrate requirements based on risks and benefits of an AI system;
  • Rewarding and encouraging responsible AI practices;
  • Leveraging takeaways from regulatory sandboxes;
  • Simplifying the Act’s oversight and enforcement; and
  • Fostering the EU’s role on AI governance at an international level.

In its response, CIPL reiterates that for the AI Act to be effective in protecting fundamental rights while also laying a foundation for a new era in EU innovation, it needs to be flexible enough to adapt to future changes. Further, CIPL argues that the Act must not be overly restrictive, so as to avoid suppressing valuable and beneficial innovations and uses of AI across a range of industries and sectors. Finally, CIPL proposes that the AI Act would benefit from targeted adjustments that better clarify the balance of responsibilities of AI providers, deployers and users, particularly for general purpose AI and open source AI models.

Read CIPL’s work on AI and data protection.