AI Act May Take a New Direction: Innovation vs. Fundamental Rights Protection
Why has the need to ease the AI Act arisen?
The EU AI Act, which entered into force in the summer of 2024, constitutes one of the most comprehensive and stringent regulatory frameworks governing artificial intelligence on a global scale. Certain provisions of the regulation are to be implemented gradually from 2025 onwards; however, already at an early stage of its implementation, significant market and political pressure – including from technology companies supported by the President of the United States – has been exerted on the European Commission with a view to relaxing the rules.
The original objective of the regulation was to place high-risk applications – in particular biometric identification, healthcare systems, and law enforcement uses – under strict regulatory control. At the same time, the AI Act does not operate as a standalone legal instrument, but forms part of the European Union’s broader digital regulatory framework, which includes, inter alia, the GDPR, the ePrivacy Directive, and national data protection regimes.
This complex regulatory environment, taken as a whole, imposes a significant compliance burden on market participants. According to the position of several industry stakeholders and Member States, the European Union must avoid overregulation in a technological field that is evolving rapidly and characterised by intense global competition. In contrast, data protection and civil society organisations have expressed concerns that any relaxation of the rules would primarily benefit large technology companies, while gradually weakening the protection afforded to users.
What has taken place during the negotiations?
In recent months, representatives of the European Parliament and the EU Member States have engaged in intensive discussions regarding potential amendments to the AI Act. According to press reports, the parties negotiated for more than twelve hours; nevertheless, no agreement was reached on the proposed modifications.
The negotiations focused on the “Digital Omnibus on AI Regulation” proposal, which aims to introduce targeted amendments in order to simplify the regulatory framework and render it more manageable for businesses, in particular by reducing administrative burdens. A key point of contention concerned whether certain sectors that are already subject to strict regulatory regimes could be exempted from specific obligations under the AI Act.
However, the debate extends beyond technical amendments: a fundamental systemic question has emerged as to the extent to which it is justified to substantially revise a regulatory framework that has already been adopted but has not yet fully entered into force. Critics argue that the proposed “simplification” could, in practice, lead to a weakening of the safeguards previously established. By contrast, the European Commission maintains that the objective of the proposed amendments is not to diminish protection, but rather to improve the practical applicability of the regulation and thereby enhance the competitiveness of the European Union.
Results of the Negotiations
Based on the current state of the negotiations, no agreement has yet been reached regarding the easing of the AI Act rules. Consequently, the final content of the regulation cannot yet be regarded as settled, and therefore the exact scope and level of strictness of certain provisions may still change.
Significant political and professional differences of opinion continue to emerge throughout the legislative process, slowing down the formation of consensus. The absence of an agreement may lead to increased regulatory uncertainty in the short term, particularly for technology developers, startups, companies operating in industrial and data-intensive sectors, as well as businesses engaged in health technology.
What Can Be Expected in the Future?
The negotiations are expected to continue in the coming months, and several possible regulatory directions are beginning to emerge. As part of a potential compromise solution, certain industries or fields of application may receive partial exemptions, while applications classified as “high-risk” would remain subject to strict regulation.
In addition, the differentiated regulatory approach may be further strengthened. In this context, the legislator may define more precisely which applications fall within the “high-risk” category, thereby contributing to the reduction of the current regulatory uncertainty and enabling more predictable operation for market participants.
At the same time, it also remains a realistic possibility that the current, stricter regulatory structure of the AI Act will remain substantially unchanged, particularly with regard to the protection of fundamental rights, data security, and transparency requirements.
Closing Remarks
The professional and political debate surrounding the AI Act extends beyond the traditional issues of technology regulation. The real significance of the regulatory process lies in how the European Union intends to strike a balance between supporting innovation, maintaining economic competitiveness, and protecting fundamental rights, as well as how effectively it can respond to the challenges posed by the rapidly evolving global technological environment.
Based on the ongoing negotiations and professional consultations, the final regulatory framework of the AI Act still cannot be considered finalized. Accordingly, it is advisable for market participants to continuously monitor the legislative process, as the expected amendments and interpretative developments may have a significant impact – both legally and commercially – on companies developing or utilizing artificial intelligence technologies.
