Background to the new legislation
As more businesses adopt artificial intelligence, the need for clear guidelines for using the technology responsibly and securely is growing. AI can offer significant benefits, but it also raises questions about transparency, data security, discrimination and accountability.
To address these challenges, the EU has adopted the AI Act, the world’s first comprehensive legislation on artificial intelligence. Its purpose is to establish common rules for the development and use of AI, allowing the technology to be used in ways that protect citizens and businesses without hindering innovation.
The legislation was adopted in 2024 and is being implemented gradually. The majority of the rules will take effect on 2 August 2026, when enforcement will also begin. Among the first requirements are increased transparency, including informing users when they are interacting with AI and ensuring that AI-generated content can be identified as artificially created.
Requirements of the EU AI Act
The AI Act introduces a range of requirements designed to ensure greater transparency and responsible use of artificial intelligence in practice. The first rules will take effect on 2 August 2026 and include stricter requirements for how AI is used in products and digital solutions.
A key focus is transparency. Users must be clearly informed when interacting with an AI solution such as a chatbot, so that it is not presented as human communication. In addition, AI-generated content such as images, audio and video must be identifiable as artificially created, including deepfakes.
The requirements for high-risk AI systems have been postponed as part of the EU’s efforts to simplify the legislation under the so-called “AI Omnibus”, which reached political agreement in May 2026. Under the current timeline, high-risk systems in areas such as recruitment, credit assessment and critical infrastructure must comply with the requirements from December 2027, while AI embedded in regulated products will be covered from August 2028.
Overall, the legislation means that businesses must consider not only how AI is developed, but also how it is used and documented in practice.
So what do you do now?
For many businesses, this means that AI is no longer just a technological or business issue. It is also a matter of governance, documentation and accountability. That is why it can be beneficial to start mapping where AI is used across the organisation’s products and processes, and which requirements may become relevant in the future.
Although not all parts of the AI Act have yet come into effect, there is good reason to start preparing now. Organisations that take a proactive approach to AI governance are often better positioned when new requirements need to be translated into practice.
We are hosting a webinar on the practical implications of the EU AI Act
If you are unsure how to translate the legislation into concrete workflows, processes and decisions, Dwarf and TheRightW.ai are hosting a webinar on 25 June from 9:00–10:00 focused specifically on this topic. We will share insights and perspectives on how businesses can approach AI governance in practice and which areas are particularly important to consider already.
The webinar is aimed at decision-makers, product owners and digital teams who want a practical understanding of how the AI Act may affect their products and organisation.
