Views: 873
Two most vital legal acts in the European digital transition and AI (i.e. the full text of AI “basic law” and the Digital Services Act) are called in the EU-jargon “digital omnibus”. These twin acts are expected to be implemented by the end of 2025; however, some officials and digital experts warned that the digital omnibus and the associated Code of Practice need amendments and corrections.
Background
The waves of critics and complaints around soon expected the digital services’ and the AI laws’ implementation are storming from all sides: even the European Commission proposed recently some actions to simplify legal texts affecting the digital rules and the sector, in general, before the acts’ publication in December. As is known, the digital omnibus covers two recently approved legislative text: i.e. a) AI Act, and b) the Digital Services Act (DSA).
Thus, the Danish digital affairs minister Caroline Stage Olsen told the Euronews that the “omnibus” needs fundamental re-shuffling. Besides, the European experts and the member states officials (e.g. the mentioned Danish digital minister) underlined, that the EU-wide digital regulations should be simplified on all fronts.
The Danish opinion is actually vital as the country during present six months, i.e. up to the end of 2025, is heading the rotating Presidency in the EU Council where the digital ministers are regularly meeting to find compromises on legislation including all the digital files.
More on Danish Presidency in: https://www.integrin.dk/2025/07/08/danish-presidency-in-the-council-some-aspects-in-the-work-plan/
Besides, the EU Technology Commissioner H. Virkkunen already announced that the Commission is carrying out a digital fitness check, which will result in the “omnibus” simplification package set to be presented on 10 December 2025.
More in the Commission’s staff working document on “fitness check” on digital fairness, in: https://commission.europa.eu/document/707d7404-78e5-4aef-acfa-82b4cf639f55_en
The Commission aims to identify reporting obligations in existing digital legislation that can be cut to ease pressure on enterprises, particularly SMEs. The question remains, notes the Euronews, whether that simplification package will cover three pertinent legal acts in the digital sector -i.e. the Digital Markets Act (DMA), mentioned above the DSA and the AI Act; these rules entered partially into force less than two years ago, but haven’t yet fully come into effect.
Source: https://www.euronews.com/next/2025/07/16/eu-should-simplify-its-ai-act-and-digital-rulebook-says-danish-minister
The EU Commission published practical guidelines on the scope of obligations for providers of general-purpose AI models under the AI legislation: specially, covering issues concerning “when an AI model becomes a general-purpose AI model”: i.e. AI model is considered a general-purpose AI model if it was “trained using an amount of computational resources that exceeds 10^23 floating point operations and if it can generate language, text-to-image or text-to-video. The guidelines outline the concepts of a ‘provider’ and of ‘placing on the market’ and clarify when an actor modifying a general-purpose AI model is considered to become a provider.
Source and citations from: https://digital-strategy.ec.europa.eu/en/news/learn-more-about-guidelines-providers-general-purpose-ai-models
Possible amendments to the AI legislation
The Danish digital minister said that it’s important to consider where the AI Act “may be imposing unnecessary burdens on companies”. Basically, the European AI law aims to regulate artificial intelligence systems according to the risk they pose to society (this clausal entered into force in August 2024, but the whole act will fully apply only in 2027).
Apparently, there are certain legal aspects in the AI legislation with a “room for simplification, or for making interpretations easier for companies, or even to make fewer reporting demands”, noted the Danish minister, adding that “it is better that companies do not spend a lot of money on unnecessary implementation”.
It is vital to mention that voluntary guidelines for providers of general-purpose AI models, such as ChatGPT and Claude AI models included in the AI Act in the form of Code of Practice (published this July), have been for long subject to lobbying efforts as companies tried to get the “simplest possible rules”.
Specific objectives of the Code of Practice:
1. To serve as a guiding document for demonstrating compliance with the obligations provided under the AI Act (art. 53 and 55), while recognising that adherence to the Code does not constitute conclusive evidence of compliance with the obligations under the AI legislation.
2. To ensure that providers of general-purpose AI models comply with their obligations under the AI Act and to enable the AI Office to assess compliance of providers of general-purpose AI models who choose to rely on the Code to demonstrate compliance with their obligations under the AI Act.
Reference to: https://code-of-practice.ai/?section=summary
Comments on the Code of Practice
The Code of Practice was developed through an extensive multi-stakeholder process launched in July 2024, involving over a thousand participants from industry, academia, civil society, rights-holders and the EU member states represented in the AI Board. 13 independent experts appointed by the AI Office led the drafting process, developed the Code through three rounds of consultations. The Code does not impose obligations; it serves as guidance to help providers meet their existing obligations under the AI Act without creating new ones, extending existing ones, or imposing additional burdens.
Starting from 2 August 2025, providers placing general-purpose AI models on the market must comply with their respective AI Act obligations. Providers must notify the AI Office of general-purpose AI models with systemic risk to be placed on the EU market without delay. In the first year – from 2 August 2025- the AI Office will offer to collaborate closely in particular with providers who adhere to the Code to ensure that models can continue to be placed on the EU market without delays.
In particular, if these providers do not fully implement all commitments immediately after signing the Code, the AI Office will not consider them to have broken their commitments under the Code and will not reproach them for violating the AI Act. Instead, in such cases, the AI Office will consider them to act in good faith and will be ready to collaborate to find ways to ensure full compliance. However, from 2 August 2026 onwards, the Commission will enforce full compliance with all obligations for providers of general-purpose AI models with fines.
More on Code in: https://digital-strategy.ec.europa.eu/en/policies/contents-code-gpai
Basically, the Code is “transforming the AI Act’s high-level requirements into concrete standards that will likely shift frontier AI companies’ practices toward safer ones”.
Source: https://aifrontiersmedia.substack.com/p/how-the-eus-code-of-practice-advances
Earlier this July, the CEOs from more than 40 European companies including ASML, Philips, Siemens and Mistral, asked for a “two-year clock-stop” on the AI Act before key obligations enter into force this August, claiming they have only a short time-frame to comply.
More on complains and the template (pattern for process) in: https://www.integrin.dk/2025/07/16/controversies-over-the-european-digital-transition-moves-to-slow-down-the-ai-law/
In an effort to boost interest in the Code, the Commission has informally recognized a one-year grace period within which “the AI Office will not consider them to have broken their commitments under the Code and will not reproach them for violating the AI Act. Instead, in such cases, the AI Office will consider them to act in good faith and will be ready to collaborate to find ways to ensure full compliance.”
More in: https://digital-strategy.ec.europa.eu/en/faqs/questions-and-answers-code-practice-general-purpose-ai
In addition, the US administration, as well as American big tech companies, expressed criticism saying the rules hinder innovation. However, as the Danish minister said that the idea “of reviewing the AI Act had nothing to do with the US position”. And added that, notes the Euronews, the European comment is about the EU “ability to stand on our own feet”.
However, indirectly, of course, it has something to do with the US, because the EU “has to keep up with the US, if it wants to lead in any technology race”.
Then, the Danish minister noted that the EU “need to wake up from our sleep and start doing something; as we need to simplify the rules for our own sake, not for others” adding that “everyone that operates in Europe needs to play by the European rules”.
On July 17, the Danish minister informed the lawmakers in the European Parliament about Danish Council Presidency’s priorities for this half year and said protecting minors online is on top of the list. “I will use the presidency to put this on top of the agenda and set a clear, political ambition that can shape EU policy in the years to come,” she said, adding that Denmark seeks to make age verification tools mandatory.
At the same time, another Danish Minister for Culture, Jakob Engel-Schmidt, announced that he aims to get support across Europe to introduce stronger protection against deepfakes after Denmark adopted a bill that would make it illegal to share digital imitations of people’s personal characteristics.
Source and citations from: https://www.euronews.com/next/2025/07/16/eu-should-simplify-its-ai-act-and-digital-rulebook-says-danish-minister (16.07)
Protecting digital safety
Commission noticed that the Code of Practice would advance AI safety: e.g. H. Papadatos, Managing Director of SaferAI, argued in AI Frontiers*) that the EU’s Code of Practice provided strong incentives for frontier AI developers to adopt measurably safer practices.
Companies following the Code gain “presumption of conformity” with AI (acts. 53 and 55), meaning regulators assume compliance if they meet the Code’s standards. This powerful incentive previously compelled widespread adoption of the EU’s Code of Practice on Disinformation.
*) See more in: https://aifrontiersmedia.substack.com/p/how-the-eus-code-of-practice-advances
The Code mandates comprehensive risk management processes requiring companies to predetermine acceptable risk levels and maintain risks below these thresholds through assessment and mitigation.
Compliance must be documented in two key instruments: a Framework (similar to existing Frontier AI Safety Policies) and a Model Report showing application of the Framework for each model (like model cards). Companies must consider specific risk categories: CBRN, cyber, loss of control, and manipulation, marking significant improvement as no current company frameworks comprehensively address all these areas.
Source: https://artificialintelligenceact.substack.com/
Finally, the Code represents substantial progress towards safer AI development, meaningfully improving current industry practices through explicit risk modelling, operationalised external evaluations and mandatory public transparency. Besides, the Code’s influence will likely extend globally, providing a regulatory blueprint worldwide.