The EU AI Act: What you need to know for 2024
We summarized the most important information about the EU AI Act and how your organization can prepare for it below. As the AI Act is in its final stages, organizations are expected to meet the first compliance requirements of the Act already in 2024. It is more important than ever to start preparing now.
What is the EU AI Act?
The EU AI Act is the world's first legal framework designed to regulate Artificial Intelligence across the EU. It aims to ensure that AI systems are safe, respect existing laws on fundamental rights, and align with the EU's values.
The AI Act introduces a risk-based approach, distinguishing between four kinds of AI systems:
1. Unacceptable
AI applications that are incompatible with EU values and fundamental rights. They will be prohibited.
2. High Risk
Highly regulated AI Systems that could cause significant harm if they are failing or misused, or that are safety components.
3. Limited Risk
Applications that pose a risk of manipulation or deceit. They are less regulated, but have transparency obligations.
4. Minimal Risk
All remaining AI systems. While they have mandatory requirements, transparency and ethical use are encouraged.
Unsure how you're impacted?
Learn how to cope with the EU AI Act with us in a 15-min call.
Together, we will assess how your organization is affected by the EU AI Act and how your AI system classifies.
trail also offers an AI Governance copilot that helps you in preparing for the AI Act.
trail is supported by leading organizations
What failing to comply with the EU AI Act means
Non-compliance with the requirements of the EU AI Act will result in high penalties:
Up to €15 million or 3% of your global annual revenue for non-compliance with any other regulatory requirements, such as those of high-risk systems or GPAI models.
Up to €7.5 million or 1.5% of your global annual revenue for providing incorrect, incomplete or misleading information about your AI systems.
When does the EU AI Act become enforceable?
The AI Act is currently in its final stage and is expected to enter into force in early 2024. After 6 months, organizations that are putting AI systems into the EU market need to comply with the regulation on prohibited practices. After 12 months, obligations for general purpose AI models (GPAI) become applicable. All other obligations, including those on high-risk AI systems, become applicable after 24 months.
Organizations are expected to meet the first compliance requirements of the EU AI Act at the end of 2024.
What high-risk system providers need to do
Especially high-risk AI system providers, such as in finance or medicine, need to fulfil strict requirements to demonstrate the trustworthiness of their systems.
Quality & Risk Management
Includes risk mitigation and model tests along lifecycle, data governance, keeping detailed documentation and keeping logs.
Conformity Assessment
Undergo self-assessment and third-party audits before putting the AI system on the market.
Registration in EU Database
High-risk system providers need to provide information about their applications, accessible to the public.
[High-risk systems] must also be traceable and auditable, ensuring that appropriate documentation is kept, including of the data used to train the algorithm that would be key in ex post investigations.”
Prepared for the EU AI Act
It takes time to meet the high requirements: traceable AI development and documentation, robust risk management and technical audits. Gain a head start on the EU AI Act by preparing now.
With trail you can avoid reputational loss and penalties of up to €35m or 7% of your annual revenue.
This is how trail's AI Governance copilot helps you to comply with the EU AI Act:
Integrate AI Act into your process
Adapt the AI Act through documentation templates to your organization's workflows.
Translate requirements
Transfer the regulatory requirements of the EU AI Act into actionable steps already during AI development.
Get pre-audit assessment
Assess your risks and governance measures for each AI project before official audit.
