Sandboxing the AI Act: testing the AI Act proposal with Europe's future unicorns

Foreword

DIGITALEUROPE is delighted to present this report from our pre-regulatory sandboxing initiative, which aimed to evaluate the proposed AI Act and its potential implications for European start-ups and SMEs.

In our manifesto for the current legislative term, A stronger digital Europe, we emphasised the need for agile and mission-based policies. We stressed the importance of clear policy goals, agile policy-making processes, and multi-stakeholder engagement through regulatory sandboxing. The outcomes presented in this report align with our commitment.

Participants in the sandbox appear genuinely committed and positive to the need for AI regulation and appreciate the clarity of rules and not least the need for applicable standards in the market. The main worries are a slow-down of European innovation, investment and market share for EU companies and a gap in operational interpretation and understanding of risk categories. Further, liability exposure has created challenges and the cost of compliance remains a significant concern, but its full extent cannot be accurately determined at this stage.

Another notable issue highlighted is the lack of existing standards, and consequently uncertainty about what formal harmonised standards will ensure compliance. Participants also show that achieving compliance will be a time-consuming process, which for many will exceed 12 months after the finalisation of technical standards.

Considering the global nature of the AI industry, the impact of the AI Act on international competitiveness is a pressing concern. Some worry that the implementation of the AI Act may hinder their ability to work with high-risk AI models in Europe and deter international investors.

The Sandboxing exercise has revealed discrepancies in compliance distribution across EU member states. Participants perceive that countries with robust governance and regulatory systems may be better equipped to manage compliance and harmonisation.

This sandboxing exercise underscores the importance of incorporating sandboxing into the fabric of the EU’s better regulation principles. By engaging in agile policy-making processes, we can boost responsible AI deployment through private public collaboration, while ensuring Europe remains at the forefront.

I extend my gratitude to all the participants who contributed their insights, as well as the dedicated teams involved in organising this sandboxing initiative. Together, we can shape a future where Europe’s digital landscape thrives, driving economic growth and societal benefits while upholding the highest standards of ethics and accountability.