The European Union’s European Artificial Intelligence Act (AI Act) has officially come into force, marking a historic moment as the world’s first comprehensive regulation on artificial intelligence. This landmark legislation sets the stage for a new era in AI governance, aiming to balance innovation with stringent safeguards to protect citizens’ rights and ensure the trustworthy use of AI technology.

Key Provisions of the AI Act

Minimal Risk AI Systems: No Obligations Required

The AI Act introduces a tiered approach to AI regulation based on risk levels. AI systems classified as minimal risk, such as AI-enabled recommender systems and spam filters, face no mandatory obligations. These systems are deemed to pose little to no risk to individuals’ rights and safety. While not required, companies may still voluntarily adopt additional codes of conduct to ensure best practices.

Transparency Requirements for Specific AI Systems

AI systems with specific transparency risks, including chatbots and AI-generated content, must adhere to new disclosure requirements. For instance, chatbots are required to inform users that they are interacting with a machine. Additionally, any AI-generated content, such as deep fakes, must be labelled to indicate its artificial origin. This aims to enhance transparency and help users make informed decisions.

High-Risk AI Systems: Stringent Compliance Measures

AI systems deemed high-risk, such as those used in recruitment or credit scoring, are subject to rigorous compliance requirements. These include implementing risk-mitigation strategies, ensuring high-quality data sets, maintaining detailed documentation, and providing robust cybersecurity measures. High-risk systems must also have mechanisms for human oversight and clear user information. Regulatory sandboxes will support innovation while ensuring adherence to these standards.

Bans on Unacceptable Risk AI Systems

The AI Act also identifies and bans AI systems considered to pose an unacceptable risk to fundamental rights. This includes applications that manipulate human behaviour or facilitate social scoring by governments or corporations. Certain uses of biometric systems, like emotion recognition in workplaces or real-time remote biometric identification in public spaces, are prohibited under strict conditions.

General-Purpose AI Models and Transparency Measures

The regulation addresses the complexities of general-purpose AI models, which perform a wide range of tasks. To manage systemic risks associated with these highly capable models, the AI Act introduces transparency requirements along the value chain. Providers must conduct self-assessments, mitigate risks, and report serious incidents to ensure these models are used responsibly.

Implementation and Enforcement

National Authorities and European AI Office

EU Member States have until 2 August 2025 to designate national authorities responsible for enforcing the AI Act and conducting market surveillance. The European Commission’s AI Office will oversee the Act’s implementation at the EU level, supported by three advisory bodies: the European Artificial Intelligence Board, a scientific panel of independent experts, and an advisory forum of diverse stakeholders.

Penalties for Non-Compliance

Firms that fail to comply with the AI Act face substantial fines. Violations related to banned AI applications can result in penalties of up to 7% of a company’s global annual turnover. Other infractions may incur fines up to 3%, and supplying incorrect information could lead to fines of up to 1.5%.

Transition and Future Steps

The majority of the AI Act’s rules will be enforced from 2 August 2026, with prohibitions on unacceptable risk AI systems starting six months earlier and regulations for general-purpose AI models taking effect after 12 months. To bridge the transitional period, the European Commission has launched the AI Pact, encouraging AI developers to voluntarily comply with key obligations before the official deadlines.

Ongoing Development and Guidelines

The Commission is developing detailed implementation guidelines and co-regulatory instruments such as standards and codes of practice. A call for expressions of interest has been issued to draft the first general-purpose AI Code of Practice, with stakeholder consultations to follow.

Expert Reactions

Margrethe Vestager, Executive Vice-President for a Europe Fit for the Digital Age, emphasised, “AI has the potential to transform our daily lives and offers significant benefits. The AI Act is a crucial step to ensure AI technology respects EU rules and safeguards fundamental rights.”

Thierry Breton, Commissioner for Internal Market, remarked, “Today marks a pivotal moment for Europe’s leadership in AI. With the AI Act in effect, Europe sets a global benchmark for ethical AI regulation and fosters a supportive environment for innovation.”

A New Paradigm for AI Regulation

The European Artificial Intelligence Act stands as the first comprehensive legal framework for AI worldwide. It aims to ensure that AI systems uphold fundamental rights, safety, and ethical principles while fostering innovation and investment across the EU. As AI technology continues to evolve, this regulation represents a critical milestone in global AI governance, positioning Europe as a leader in the responsible development and deployment of AI technologies.

  AI Act, AI transparency, EU AI regulation, European Artificial Intelligence Act, general-purpose AI models, high-risk AI systems, Margrethe Vestager, Thierry Breton