📢 KEY PROVISIONS OF THE EU AI ACT APPLICABLE FROM 2 FEBRUARY 2025

The EU Artificial Intelligence Act ("EU AI Act"), the first comprehensive regulatory framework for Artificial Intelligence in Europe, is entering its first phase of applicability. From 2nd of February 2025, Chapter I and Chapter II will take effect.

Chapter 1 | General Provisions (Articles 1-4)

This chapter establishes the foundation for AI governance in the EU, including its scope, legal definitions, and the introduction of AI literacy as a crucial aspect of compliance.

Article 1: Subject Matter & Scope

• Defines the AI Act’s objectives: ensuring safe, ethical, and human-centric AI while fostering innovation

• Applies to AI providers, deployers, and users within the EU and beyond if their AI impacts EU citizens

Article 2: Scope

• Covers AI systems placed on the EU market or affecting individuals in the EU, regardless of where they are developed

• Excludes military AI and AI research that is non-commercial

Article 3: Definitions

• Defines AI systems, high-risk AI, deployers, and providers

• Clarifies AI as a machine-based system that generates outputs such as predictions, recommendations, or decisions

Article 4: AI Literacy

• Recognises AI literacy as a fundamental element for businesses, regulators, including their staff as well as for individuals

• Encourages AI education and awareness to enable responsible use and decision-making

• Promotes public and industry training on AI risks, capabilities, and limitations

Chapter 2 | Prohibited AI Practices (Article 5)

The AI Act bans certain AI applications that pose unacceptable risks to fundamental rights and democracy, including:

• Subliminal manipulation – AI that influences behaviour in ways that cause harm

• Exploitation of vulnerabilities – AI systems that take advantage of individuals based on age, disability, or socio-economic factors

• Real-time biometric surveillance (public spaces) – Strictly banned, except under limited law enforcement conditions with prior authorisatio

• Social scoring AI – AI that ranks individuals and leads to discriminatory treatment

• Predictive policing based solely on profiling – AI systems that assess criminal risk without human oversight

What’s Next? EU AI Act Implementation Timeline

The EU AI Act is being introduced in phases, with different obligations taking effect over the coming years:

• 2 February 2025 – Prohibited AI practices become enforceable

• 2 August 2025 – AI governance structures begin taking shape

• 2 February 2026 – High-risk AI obligations come into force

• 2 August 2026 – AI regulatory sandboxes must be operational

• 2 August 2027 – General-purpose AI model compliance deadline

• 31 December 2030 – Final compliance deadline for existing high-risk AI systems

How should businesses prepare?

With enforcement deadlines approaching, businesses must act now to ensure compliance and mitigate risks.

Evaluate AI Models & Compliance Gaps

• Conduct an AI risk audit – classify AI systems as minimal, limited, high-risk, or prohibited

• Ensure transparency & documentation – maintain records, explainability, and data governance

• Prepare for third-party audits if deploying high-risk AI

Engage with Regulators & Industry Experts

• Monitor EU AI guidance and sector-specific standards

• Join regulatory sandboxes (from August 2026) to test AI models under compliance oversight

• Collaborate with legal professionals & compliance bodies for tailored risk strategies

Invest in AI Literacy & Governance

• Train teams on AI ethics, risk management, and regulatory requirements

• Appoint an AI compliance officer and implement internal audits.

• Strengthen cybersecurity & data protection to align with GDPR and AI transparency rules

Act now to future-proof your AI systems and stay ahead of the EU AI Act.

What Should Businesses Do Now?

Organisations developing or deploying AI should:

• Assess AI risks and compliance readiness

• Train teams on AI literacy to meet Article 4 requirements.

• Prepare for upcoming obligations in 2026 and 2027

For further legal guidance on the EU AI Act? Contact info@pelaghiaslaw.com to ensure your AI systems align with compliance requirements.