Understanding the Implications of the AI Act for Businesses

Risk & Governance Legal Newsletter (February 2025)

On July 12, 2024, the EU officially published Regulation (EU) 2024/1689, known as the Artificial Intelligence Act (the “AI Act”). This regulation marks the world’s first comprehensive legal framework governing AI.

Discussions on the AI Act had been ongoing since the European Commission initially proposed the legislation on April 21, 2021. The rapid advancement of generative AI and other technologies necessitated continuous updates to the proposal. Ultimately, the AI Act was formally adopted after receiving approval from the Council of the EU on May 21, 2024.

The AI Act entered into force on August 1, 2024, with most of its provisions set to take effect from August 2, 2026. Certain provisions, however, will be enforced at different times as appropriate. The AI Act applies not only to companies within the EU but also broadly to businesses providing AI systems and related services within the EU. As a result, businesses outside the EU cannot overlook the implications of this regulation.

1. Step 1: Identification of AI System

The AI Act primarily regulates AI Systems and general-purpose AI models. Their respective definitions are as follows:

AI System (Article 3(1)): A machine-based system that is designed to operate with varying levels of autonomy and may exhibit adaptiveness after deployment, inferring outputs such as predictions, content, recommendations, or decisions from the input it receives.

General-purpose AI model (Article 3(63)): An AI model capable of competently performing a wide range of tasks, trained with a large amount of data using self-supervision at scale, except for AI models used for research, development, or prototyping activities before they are placed on the market.

2. Step 2: Identification of Regulated Persons

The AI Act applies to various stakeholders involved with AI systems. The primary regulated parties are:

  • Providers: Entities placing AI systems on the market or putting them into service in the EU, regardless of their location.
  • Deployers: Entities using AI systems within the EU.

Businesses must identify which category of regulated parties they fall under and verify where the AI systems are being developed, placed on the market, or used.

3. Step 3: Exemption Eligibility

The AI Act provides exemption provisions for certain AI systems. If a system qualifies for an exemption, it will not be subject to regulation. Exemptions primarily include:

  • AI systems used exclusively for military, defense, or national security purposes.
  • AI systems developed for scientific research and development.
  • Deployers who are natural persons using AI systems in a personal non-professional capacity.

4. Step 4: Categorization of AI Systems

The AI Act adopts a risk-based approach, categorizing AI systems based on the risks they pose:

  • Prohibited AI Practice: Systems that could cause significant harm to individuals.
  • High-risk AI System: Regulated systems that require compliance obligations.
  • AI System with Limited Risk: Systems subject to transparency obligations.
  • Other AI Systems: Not subject to regulation under the AI Act.

5. Step 5: Identification of Applicable Regulations

Operators need to take different approaches based on the categories classified in Step 4:

  • Prohibited AI Practice: Banned from being placed on the market, subject to penalties.
  • High-risk AI System: Must comply with several obligations, including risk management and data governance.
  • AI System with Limited Risk: Subject to transparency obligations.

6. Step 6: Preparations for Implementation

The AI Act came into effect on August 1, 2024, with provisions enforcing sequentially. Key dates include:

  • Chapter 1 and Chapter 2: February 1, 2025.
  • High-risk AI system: August 2, 2027.
  • General-purpose AI model: August 2, 2025.

In conclusion, the AI Act establishes a framework intended to ensure that AI technologies are developed and used safely and responsibly within the EU, while also impacting businesses worldwide. It is crucial for companies to understand their obligations under this new regulation to ensure compliance and safeguard their operations.

More Insights

Enhancing AI Safety through Responsible Alignment

The post discusses the development of phi-3-mini in alignment with Microsoft's responsible AI principles, focusing on safety measures such as post-training safety alignment and red-teaming. It...

Mastering Sovereign AI Clouds in Intelligent Manufacturing

Sovereign AI clouds provide essential control and compliance for manufacturers, ensuring that their proprietary data remains secure and localized. As the demand for AI-driven solutions grows, managed...

Empowering Ethical AI in Scotland

The Scottish AI Alliance has released its 2024/2025 Impact Report, showcasing significant progress in promoting ethical and inclusive artificial intelligence across Scotland. The report highlights...

EU AI Act: Embrace Compliance and Prepare for Change

The recent announcement from the EU Commission confirming that there will be no delay to the EU AI Act has sparked significant reactions, with many claiming both failure and victory. Companies are...

Exploring Trustworthiness in Large Language Models Under the EU AI Act

This systematic mapping study evaluates the trustworthiness of large language models (LLMs) in the context of the EU AI Act, highlighting their capabilities and the challenges they face. The research...

EU AI Act Faces Growing Calls for Delay Amid Industry Concerns

The EU has rejected calls for a pause in the implementation of the AI Act, maintaining its original timeline despite pressure from various companies and countries. Swedish Prime Minister Ulf...

Tightening AI Controls: Impacts on Tech Stocks and Data Centers

The Trump administration is preparing to introduce new restrictions on AI chip exports to Malaysia and Thailand to prevent advanced processors from reaching China. These regulations could create...

AI and Data Governance: Building a Trustworthy Future

AI governance and data governance are critical for ensuring ethical and reliable AI solutions in modern enterprises. These frameworks help organizations manage data quality, transparency, and...

BRICS Calls for UN Leadership in AI Regulation

In a significant move, BRICS nations have urged the United Nations to take the lead in establishing global regulations for artificial intelligence (AI). This initiative highlights the growing...