Landmark Regulation Takes First Steps

The European Union’s Artificial Intelligence Act, the world’s first comprehensive AI regulation framework, began its phased implementation on 2 February 2025. This landmark legislation represents a watershed moment for technology regulation, establishing clear rules for AI development and deployment across all 27 EU member states.

As the first provisions come into effect, businesses operating in or serving EU markets must now navigate new compliance requirements designed to ensure AI systems are developed responsibly while safeguarding fundamental rights including privacy, non-discrimination, and fairness.

Impact on Businesses

The EU AI Act introduces a risk-based approach to AI regulation, with different obligations based on the potential harm an AI system could cause:

For businesses deploying or developing AI systems, the immediate implications include:

• Immediate bans on “unacceptable risk” AI systems that manipulate human behaviour, exploit vulnerabilities, or enable social scoring by public authorities
• AI literacy requirements mandating that companies ensure their staff understand AI systems through appropriate policies, procedures, and training
• Transparency obligations for certain AI applications, requiring clear disclosure when customers interact with AI systems
• Data quality and governance standards to ensure AI systems are trained on appropriate, representative datasets
• Regular risk assessments to identify and mitigate potential harms from AI deployment

Technology companies, particularly those developing or deploying high-risk AI systems in areas such as healthcare, education, employment, and critical infrastructure, will face the most significant compliance burdens as the regulation continues to roll out.

Upcoming Implementation Phases

The EU AI Act implementation timeline continues with several key milestones:

• May 2025: Introduction of Codes of Practice for general-purpose AI models
• August 2025: Transparency requirements, governance rules for general-purpose AI providers, and notification requirements to authorities become applicable
• August 2026: End of the 24-month transition period, with obligations for high-risk AI systems taking effect
• August 2027: Complete implementation of the Act, including obligations for high-risk AI systems as safety components

Non-compliance penalties can reach up to €35 million or 7% of total worldwide annual turnover, making preparation essential for businesses of all sizes.

Key Business Requirements

As the EU AI Act continues its phased implementation, businesses will need to address several critical requirements:

• AI System Classification: Companies must determine whether their AI systems fall into prohibited, high-risk, limited-risk, or minimal-risk categories, with each category carrying different obligations
• Risk Management Systems: High-risk AI providers must implement robust risk management processes throughout the entire AI lifecycle
• Technical Documentation: Maintaining comprehensive documentation detailing system architecture, training methodologies, and validation processes
• Human Oversight: Ensuring appropriate human supervision for high-risk AI systems
• Data Governance: Implementing rigorous data quality standards for training, validation, and testing datasets
• Transparency Requirements: Clearly informing users when they interact with AI systems
• Cybersecurity Measures: Establishing appropriate security protocols to protect AI systems from vulnerabilities
• Registration: High-risk AI systems must be registered in an EU database before being placed on the market
• Conformity Assessment: Conducting internal or external assessments to verify compliance
• Post-Market Monitoring: Establishing systems to identify and address issues after deployment

How Wolja Digital Can Help

At Wolja Digital, our team of legal and technical experts can guide technology startups through the complexities of these new regulatory requirements. We provide strategic advice on compliance approaches that align with business objectives while meeting the EU AI Act’s standards.

Contact us to learn more about navigating this new regulatory landscape.