Enterprise AI Analysis
Do Citizens Agree with the EU AI Act? Public Perspectives on Risk and Regulation of AI Systems
Gabriel Lima
Max Planck Institute for Security and Privacy, Bochum, Germany
Frederike Zufall
Department of Informatics, Karlsruhe Institute of Technology, Karlsruhe, Germany
Waseda Institute for Advanced Study, Waseda University, Tokyo, Japan
Gustavo Gil Gasiola
Department of Informatics, Karlsruhe Institute of Technology, Karlsruhe, Germany
Yixin Zou
Max Planck Institute for Security and Privacy, Bochum, Germany
The European Union (EU) has spearheaded the regulation of artificial intelligence (AI) with the AI Act, which regulates AI systems based on the risks they pose to fundamental rights and other protected values. AI systems that pose unacceptable risks are prohibited, high-risk AI systems must comply with mandatory requirements, and minimal risk AI systems are encouraged—but not required—to adopt voluntary standards. Motivated by concerns that the AI Act may not reflect the public's opinions, we investigate how laypeople (N=1,421) assess 48 different AI systems concerning their risk and regulation. We find that people believe all 48 AI systems pose moderate levels of risk and should be regulated (albeit without outright prohibitions). Our findings challenge the AI Act's tiered approach, showing that people might support horizontal regulation requiring minimal standards for AI systems, and provide implications for developers seeking to develop AI aligned with public expectations.
Key Executive Impact
This study offers crucial insights for leaders navigating AI regulation and public perception. Understand the core findings that challenge current AI governance models and inform future strategy.
1,421+
Participants Surveyed
48
AI Systems Assessed
4
Countries Represented
100%
AI Systems Perceived as Risky
Deep Analysis & Enterprise Applications
Select a topic to dive deeper, then explore the specific findings from the research, rebuilt as interactive, enterprise-focused modules.
Our Research Methodology
We conducted a multi-country large-scale study (N=1,421) to capture laypeople's perceptions of the risk posed by 48 different AI systems, as well as to collect their opinions about how each AI system should be regulated. This section outlines the rigorous process to ensure robust and reliable data collection.
Enterprise Process Flow
Read Definitions & Comprehension Check
→
Evaluate 4 Random AI Systems (Risk & Regulation)
→
Rate Importance of AI Act Requirements (Last AI System)
→
Demographics & Attention Check
Public Perception of AI Risk
Contrary to the AI Act's tiered approach, citizens perceive all AI systems, regardless of their official risk classification, as posing moderate levels of risk. This indicates a potential gap in public understanding or a demand for more comprehensive safety standards across the board.
Moderate Risk
All 48 AI Systems Perceived as Moderately Risky, Challenging the AI Act's Tiered Approach.
Case Study: Generalized Risk Perception
In our study, participants consistently rated all AI systems, from minimal to unacceptable risk categories, as posing moderate risk. For instance, even minimal-risk systems were rated above 'no risk at all'. This suggests a public expectation for basic safeguards across all AI deployments.
Implication for Enterprises: Companies cannot assume low-risk categorization by the AI Act equates to low public concern. Proactive measures for perceived "minimal" risk AI systems can build trust and acceptance.
Alignment with AI Act Regulation
While people support regulation for all AI systems, they largely oppose outright prohibitions. This finding suggests a preference for robust oversight and mandatory standards rather than blanket bans, even for systems deemed 'unacceptable' by the AI Act.
| Aspect | EU AI Act Approach | Public Perspective |
|---|---|---|
| Risk Categorization |
|
|
| Prohibition vs. Regulation |
|
|
Future of AI Development & Governance
The consistent views across EU countries and the US suggest a potential for unified international AI governance. For developers, this implies that aligning with AI Act requirements, even for minimal risk systems, can foster public trust and broader adoption.
Global Alignment
Similar AI Risk Perceptions Across EU and US Indicate Potential for Unified International AI Governance.
Case Study: Cross-Jurisdictional Consensus
Our study observed surprisingly minimal differences in AI risk perception and regulatory preferences across Germany, Spain, France, and the United States. This suggests a foundational consensus among laypeople that AI systems require thoughtful regulation, challenging the notion of fragmented global AI perspectives.
Implication for Enterprises: For global companies, this convergence hints at the feasibility and benefit of adopting a single, high-standard compliance framework, potentially based on the AI Act, to meet diverse market expectations and streamline international deployment strategies.
Calculate Your Potential AI ROI
Estimate the efficiency gains and cost savings your enterprise could achieve by strategically implementing AI solutions aligned with public expectations and robust governance frameworks.
Estimated Annual Savings
$0
Annual Hours Reclaimed
0
Your AI Implementation Roadmap
A phased approach to integrating AI, ensuring compliance, trust, and optimal performance based on global public perceptions and regulatory insights.
Phase 01: Strategic Assessment & Public Perception Audit
Conduct a comprehensive audit of existing and planned AI systems against public perception benchmarks and the AI Act's risk framework. Identify areas of misalignment and potential trust gaps.
Phase 02: Horizontal Standards & Trust-Building Framework
Develop and implement a horizontal standard framework for all AI systems, incorporating basic requirements for transparency, data quality, and human oversight, even for minimal-risk AI.
Phase 03: AI Literacy & User Empowerment Programs
Launch internal and external AI literacy initiatives to educate users on AI capabilities, limitations, and inherent risks, fostering accurate perceptions and appropriate reliance on AI systems.
Phase 04: Continuous Monitoring & Adaptive Governance
Establish a robust system for continuous monitoring of AI system performance, ethical impact, and regulatory compliance. Implement adaptive governance mechanisms to respond to evolving public expectations and technological advancements.
Ready to Align Your AI Strategy?
Don't let regulatory uncertainty or public perception hinder your AI innovation. Our experts are ready to help you build compliant, trusted, and high-performing AI systems.
Ready to Get Started?
Book Your Free Consultation.
Let's Discuss Your AI Strategy!
Lets Discuss Your Needs
Select a Date
Sun
Mon
Tue
Wed
Thu
Fri
Sat