Running a business today often means using artificial intelligence somewhere in the process. It may help analyse data, automate repetitive tasks or support customer service. Yet for many companies, one of the biggest challenges has not been the technology itself, but understanding which regulatory rules apply.
A new amendment to the EU AI Act aims to make that task easier. Published through Regulation (EU) 2026/1744, the changes simplify parts of the existing framework, reduce unnecessary administrative burdens and clarify several areas that created uncertainty during the first stages of implementation. At the same time, the legislation introduces stricter safeguards against some of the most harmful uses of generative AI.
For investors, business owners and internationally active professionals, the message is straightforward: some compliance obligations become clearer, but responsible use of AI remains just as important.
Why this matters beyond the technology sector
Many people assume AI regulation concerns only software developers. In reality, the impact is much broader. Companies across financial services, professional services, healthcare, manufacturing and retail increasingly rely on AI-powered tools to improve efficiency and decision-making.
When regulatory requirements are unclear, businesses often delay projects, allocate additional resources to legal reviews or limit the use of potentially beneficial tools. The amendment seeks to reduce those barriers by removing overlaps between the AI Act and other EU legislation where similar requirements already exist.
For investors, this matters because regulatory uncertainty can translate into higher operational costs and slower innovation. When businesses gain a clearer understanding of their obligations, they can make strategic decisions with greater confidence and focus more resources on growth and development.
Where the rules have become clearer
One of the most significant changes concerns the definition of an AI safety component.
During the initial implementation of the AI Act, questions emerged about whether certain supporting or optimisation tools could be classified as high-risk AI systems. The amendment provides additional clarification and confirms that systems performing purely supportive functions should not automatically fall into the high-risk category simply because they are used in regulated environments.
This may sound technical, but the practical effect can be substantial. A company using AI to streamline workflows or assist employees may no longer face the same uncertainty about whether extensive high-risk obligations apply.
Another notable change relates to AI literacy. Earlier requirements led many organisations to question exactly how they were expected to demonstrate employee competence in AI-related matters. The revised approach focuses less on proving a specific level of knowledge and more on taking reasonable measures to promote AI literacy among staff.
This shift recognises an important reality: effective AI governance is not achieved through box-ticking exercises. It depends on ensuring that employees understand how AI tools work, where their limitations lie and when human judgement remains essential.
Where regulation has become stricter
While the amendment reduces administrative complexity in several areas, it also strengthens restrictions on certain harmful uses of AI.
In particular, it expands prohibitions related to AI systems designed to create or manipulate intimate content without the consent of the individuals involved. The legislation also further clarifies restrictions concerning material related to child sexual abuse.
These provisions reflect growing concerns about misuse of generative AI technologies. For businesses, this serves as a reminder that innovation and responsibility must develop together. Organisations deploying AI systems will increasingly need clear internal policies regarding acceptable use, oversight and risk management.
For investors, governance should remain an important consideration when evaluating technology companies. A firm’s approach to AI-related risks can influence not only regulatory exposure but also reputation, customer trust and long-term business performance.
“The most valuable regulatory reforms are often those that remove uncertainty rather than simply remove rules.”
Simplification does not mean deregulation. The amendment reduces complexity in some areas, but organisations using AI still need to understand how their systems operate, what risks they create and which obligations continue to apply. Assuming that AI compliance requirements have disappeared could lead to costly mistakes.
What we recommend
- Review how AI is used within your business and identify which systems may be affected by the revised rules.
- When assessing investments, consider AI governance alongside financial performance and growth potential.
- Ensure employees receive practical guidance on the responsible use of AI tools, particularly where business decisions or personal data are involved.
FAQ
- Does the amendment remove the AI Act’s requirements? No. The AI Act remains in force. The amendment mainly clarifies existing provisions, reduces certain administrative burdens and addresses areas where implementation proved difficult.
- Will fewer AI systems be classified as high-risk? Not necessarily. The amendment provides clearer definitions, which may help some organisations determine that particular supporting tools do not fall within the high-risk category.
- Why should investors pay attention to AI regulation? Regulatory requirements can affect business costs, product development, operational risk and corporate reputation. All of these factors may influence a company’s long-term value.
- What changed regarding AI literacy? The revised rules place greater emphasis on promoting AI literacy through appropriate measures rather than demonstrating a predefined level of employee knowledge.
- Does this only affect large technology companies? No. Any organisation that develops, deploys or relies on AI systems may be affected, including businesses in financial services, consulting, healthcare, manufacturing and other sectors.
- Are new restrictions being introduced? Yes. The amendment expands and clarifies prohibitions relating to certain harmful applications of AI, including the creation of intimate content without consent and other serious forms of misuse.
