EU's AI Act Code of Practice Delayed Until Late 2025: What's Next for Tech Giants?

Introduction

In July 2025, the European Commission announced that a code of practice designed to assist companies in complying with the upcoming EU Artificial Intelligence Act might not be in effect until late 2025. This code, although voluntary, is a critical element as it provides legal certainty to companies that choose to sign it. Significant lobbying from major tech companies such as Google and Meta, as well as European businesses and governments, has resulted in this delay. These entities have expressed concerns over the absence of finalized guidelines, arguing that this gap could hinder compliance efforts and potentially stifle innovation within the industry .

The AI Act's guidelines for GPAI models are set to become legally binding by August 2025, and the enforcement will start a year later for new models. Existing models must comply by August 2027. The introduction of clear and structured guidelines is intended to ensure these transitions protect both consumers and businesses. By constructing a comprehensive legal framework, the European Union aims to lead the world in creating trustworthy AI systems while managing risks associated with AI applications .

The delay in the code of practice has, however, sparked a mixed reaction among industry experts and the public. While some experts view this delay as a potential opportunity for more flexible regulatory processes, others see it as a source of uncertainty that might deter investment in AI development within Europe. It is feared that the delay could blunt Europe's competitive edge at a time when AI is pivotal to technological advancements globally. Critics argue that without clear regulatory landscapes, businesses may face increased risks and costs related to compliance .

Background and Announcement

In a significant announcement, the European Commission has revealed that the much-anticipated code of practice aimed at assisting companies in adhering to the EU's Artificial Intelligence Act might not be ready until the end of 2025. This decision comes after persistent lobbying from major tech corporations like Google and Meta, which along with various European businesses and EU governments, have sought a delay in the enforcement of the AI Act. The proposed code, while voluntary, is crucial as it promises legal certainty to companies that choose to follow it. This delay has created a complex scenario where the rules for Generative Pre-trained Transformers (GPAI) models will become legally binding by August 2, 2025, with enforcement starting a year later for new models entering the market.

The background of this delay is deeply rooted in the challenges faced by businesses due to the absence of a finalized code of practice. Large technology firms argue that without these clear guidelines, their capacity to ensure compliance is severely hindered, a scenario that could stifle innovation in a rapidly evolving industry. Their concerns highlight the balance that must be struck between fostering technological advancement and maintaining robust regulatory standards. Signatories to the Code of Practice gain a degree of legal assurance, while companies that do not participate may encounter greater challenges in proving compliance with the AI Act. This voluntary code is therefore seen as an essential element in the broader regulatory framework being developed by the EU to govern AI use.

Lobbying and Delays

Lobbying from major technological players like Google and Meta has significantly influenced the timeline for implementing the European Union's Artificial Intelligence Act, illustrating the powerful role that corporate interests play in shaping legislation. These companies, along with other European businesses and governments, have argued for a delay, highlighting a lack of final guidelines that could hinder compliance if rushed. Their influence led to the European Commission's announcement that a code of practice might not be ready until the end of 2025, aligning with the concerns about stifling innovation without clear regulations .

The delay in the European Union's AI Act due to lobbying pressures underscores the challenges regulators face in balancing swift legislative action with industry needs. By postponing the code of practice, the EU aims to offer companies more time to prepare, yet the absence of these guidelines poses a period of uncertainty for businesses striving to align with upcoming rules . This gives companies the breathing space to influence the specifics of the regulation, ensuring it aligns with their operational realities.

Companies choosing not to engage with the voluntary code of practice may find themselves at a strategic disadvantage, especially since the EU has stipulated that rules for General Purpose AI models will be enforceable from August 2025, with new models required to comply a year later. Such regulations emphasize that the delay in the code's publication does not equate to a reprieve from compliance responsibilities, reinforcing the necessity for businesses to proactively adjust to future requirements .

These lobbying activities exhibit not just a conflict over regulatory timing but also reflect broader debates about the balance between fostering rapid technological advancement and maintaining ethical oversight. The ongoing discussions highlight fears that the delay may unduly benefit large U.S. tech corporations at the expense of smaller European players . Critics argue that delays in enforcement could permit the continued use of biased AI systems, underlying the ethical imperative of timely and fair regulation.

Voluntary Code of Practice

A Voluntary Code of Practice is designed to provide clarity and guidance to businesses and organizations as they navigate the complexities of emerging regulations. In Europe, the development of such a code related to artificial intelligence (AI) has been delayed until the end of 2025, according to the European Commission [0](https://www.reuters.com/business/media-telecom/code-practice-help-companies-with-ai-rules-may-come-end-2025-eu-says-2025-07-03/). While this code is not mandatory, it stands to offer substantial legal certainty to those who adhere to its guidelines, potentially reducing the risks associated with non-compliance as more binding AI regulations come into play.

In the digital realm, particularly with technologies driven by AI, the call for voluntary compliance measures stems from a desire to foster innovation while ensuring that ethical standards are upheld. Companies like Google and Meta are among those advocating for these voluntary measures in light of upcoming regulations under the EU's AI Act [0](https://www.reuters.com/business/media-telecom/code-practice-help-companies-with-ai-rules-may-come-end-2025-eu-says-2025-07-03/). This balance is crucial to maintaining a competitive edge in technology development without compromising the ethical use and deployment of AI systems.

Despite being labeled 'voluntary', the expectations for adherence to this code reflect broader strategic interests. By aligning voluntarily with the code's principles, companies may not only secure a more solid defense against future regulatory scrutiny but also enhance their credibility with consumers who are increasingly concerned with ethical practices in AI deployment. The voluntary nature, then, serves as a precursor to mandatory requirements, familiarizing companies with operational changes they might have to make in advance of stricter regulations [0](https://www.reuters.com/business/media-telecom/code-practice-help-companies-with-ai-rules-may-come-end-2025-eu-says-2025-07-03/).

The voluntary code is seen as a strategic compromise, acknowledging the rapid pace of technological advancement and the need for regulatory bodies to keep up without stifling growth. The European Commission's decision to postpone the code's implementation reflects both the pressures from lobbying by major tech firms and the EU's need to refine its regulatory approach [0](https://www.reuters.com/business/media-telecom/code-practice-help-companies-with-ai-rules-may-come-end-2025-eu-says-2025-07-03/). This delay might provide a window for additional industry input and adjustments that ensure the code not only protects user interests but also aligns with the operational realities of tech companies.

While the voluntary code of practice remains an influential tool for shaping industry standards, its delayed implementation highlights ongoing tensions between innovation and regulation. Experts suggest this period of delay could serve as an opportunity for businesses to innovate within a clearer framework once it is established, potentially setting new standards for AI ethics and business practices. This dynamic illustrates how voluntary codes, although not compulsory, become vital components in the broader regulatory landscape, helping bridge gaps between policy development and technological progress [0](https://www.reuters.com/business/media-telecom/code-practice-help-companies-with-ai-rules-may-come-end-2025-eu-says-2025-07-03/).

Enforcement Timeline for AI Rules

The European Union's timeline for enforcing artificial intelligence rules has become a central topic of discussion among businesses and policymakers alike. Recently, the European Commission announced that a comprehensive code of practice aimed at assisting companies in complying with the EU's Artificial Intelligence Act might not be ready until the end of 2025. This delay comes as a response to significant lobbying from major tech companies such as Google and Meta, alongside certain European businesses and governments. They have advocated for postponing the AI Act's implementation, citing the absence of clear guidelines as a barrier to compliance. The code of practice, while voluntary, offers legal assurances to those who choose to adopt it. However, the rules for General-Purpose Artificial Intelligence (GPAI) models will become legally binding on August 2, 2025, with a one-year grace period for new models and a two-year window for existing models to ensure compliance (source).

One of the core reasons behind the lobbying efforts by technology companies is the uncertainty surrounding the finalized code of practice. Without specific guidelines, companies argue that it's challenging to align their operations with the AI Act, potentially hindering innovation in the sector. This uncertainty extends to concerns that, without signing the voluntary code, companies may face legal ambiguities and difficulties in demonstrating compliance with AI legislation. On August 2, 2025, the binding rules for GPAI models will take effect. For models introduced after this date, enforcement will commence a year later, while existing models have until August 2, 2027, to adhere to the regulations (source).

The timeline for the AI rules is not just about dates and deadlines; it's reflective of the broader challenges facing AI governance in Europe. As major tech firms and EU governments push for delays, the debate highlights the possible friction between regulatory rigor and maintaining Europe’s competitive edge in the AI race. While some view the European Commission's delay as necessary for regulatory flexibility, others see it as creating a climate of uncertainty that could deter investment, particularly affecting startups and small to medium enterprises that lack the resources to manage complex compliance requirements. This situation underscores the delicate balance the EU must maintain between fostering innovation and ensuring accountability in AI technology (source).

Moreover, the public reactions and internal divisions within the EU highlight the complexities involved in establishing robust AI regulations. While some stakeholders argue for regulatory delays to allow for clearer guidelines and prevent stifling innovation, others are concerned about the potential for business-friendly rules to undermine the integrity of AI governance. The delay of the AI Act's enforcement timeline also raises questions about the EU's ability to effectively regulate emerging technologies and set global standards. As the grace period discussions continue, it remains to be seen how the European Commission will navigate these challenges and what compromises will ultimately shape the future of AI policy in Europe (source).

Expert Opinions on the Delay

The European Commission's decision to delay the code of practice tied to the EU's Artificial Intelligence Act has stirred diverse opinions among experts. Some industry analysts stress that the delay may lead to increased uncertainty for businesses, potentially deterring investments in AI projects within Europe. This, according to critics, could have a cascading effect, disrupting the momentum of AI innovation across the continent. The absence of clear regulatory guidelines is seen as a stumbling block that could leave companies vulnerable to unforeseen compliance challenges once the rules come into effect [].

Conversely, there are experts who perceive the delay as beneficial, offering regulatory bodies and tech companies the needed time to adapt and prepare more effectively. This perspective suggests that the postponement could ultimately lead to a more robust and adaptable regulatory framework, ensuring that the measures are well-integrated and practical to follow. By alleviating immediate compliance pressures, businesses may find room to innovate without the looming threat of stringent regulations in the short term. This flexibility might prove advantageous to European tech stocks, as it potentially reduces immediate compliance costs [].

Furthermore, the debate also highlights a fundamental concern: balancing innovation with regulation. Some experts warn that prolonged delays might send a message of regulatory indecisiveness, affecting the EU's standing as a leader in AI governance. However, others argue that even delayed, the EU's regulatory attempts are crucial in setting a precedent for ethical AI use worldwide. This duality underscores the importance of having a well-funded and coordinated approach to AI regulation that prioritizes both innovation and safety [].

Public Reactions

The announcement by the European Commission to delay the implementation of a code of practice for the EU's Artificial Intelligence Act has sparked a variety of public reactions, revealing a complex landscape of opinions and concerns. Many in the public sphere worry that delaying the code until the end of 2025 may undermine the EU's ambitious goals to be a leader in AI regulation. There is a fear that prolonged uncertainty could stifle innovation and investment within the region, particularly affecting smaller firms and startups that may lack the resources to adapt quickly [Computerworld].

Critics have also expressed apprehension towards the influence of major tech companies such as Google and Meta, which have been active in lobbying efforts against the swift implementation of the AI Act. These critics argue that the delay caters more to corporate interests, potentially weakening the regulations designed to ensure accountability and safety in AI deployment. The public seems divided, with some supporting the need for thorough and carefully planned guidelines, while others see it as a capitulation to lobbying pressures [Tech Research Online].

On the other hand, there is a contingent of the public that appreciates the complexity surrounding the AI Act's implementation. These individuals recognize that crafting comprehensive and effective regulations for such a rapidly evolving technology is a significant challenge. They believe that taking additional time to refine and adapt the code of practice may ultimately lead to stronger and more reliable legislation that better serves public interest without sacrificing innovation [Computerworld].

Furthermore, the delay has spotlighted a broader public discourse on global AI governance. The EU’s approach to regulating artificial intelligence is seen as a test case for balancing innovation with ethical considerations and public safety. Public concerns are intensifying around whether the EU's regulatory path will set a standard that others might follow or become a cumbersome model that stifles progress [VinciWorks].

Economic Impacts of the Delay

The delay in implementing the EU's Artificial Intelligence Act and its accompanying code of practice is expected to have profound economic ripple effects. On the one hand, this postponement might offer temporary relief and flexibility for companies that were scrambling to comply with stringent guidelines. However, on the other hand, the delay can create a climate of uncertainty that may deter investment, especially for European startups and small to medium-sized enterprises (SMEs). Without a clear regulatory path, firms may find it challenging to strategize and allocate resources effectively, potentially slowing innovation and growth in the region's AI sector.

The potential economic impacts extend beyond European borders, influencing international markets as well. In a competitive global landscape, the EU's decision to delay could inadvertently give technological hubs in other regions, such as the United States or East Asia, a head start in AI development and deployment. This strategic delay risks ceding ground in the race to become a leader in trustworthy AI. European companies, particularly those reliant on AI for pivotal advancements, may experience a significant drain on their competitive edge as they adapt to pending and unclear regulations.

Moreover, the forecasted increase in compliance costs due to this delay cannot be overlooked. Firms developing high-risk AI systems may find themselves navigating complex compliance landscapes without clear guidance. This can result in expenditures that are not only inflated but also potentially unnecessary, should interim measures or revisions to compliance requirements be introduced once the code is finalized. Such financial strains are likely to disproportionately impact SMEs, widening the gap between large tech entities capable of absorbing additional costs and smaller players who may struggle to remain competitive.

The economic implications of the delay are multifaceted and far-reaching, potentially impacting not just individual businesses but also broader industry trends and policy outcomes across the EU. There is a dual-edged sword in this delay: while it might provide breathing room for a more thoughtful integration of the AI Act, the uncertainty it ushers in may weigh heavily on economic optimism and investment willingness. Ultimately, how businesses and regulators address these challenges will shape the European economic landscape in the AI domain, influencing everything from job creation to the region's technological innovation trajectory.

Social Impacts and Trust Issues

The ongoing delay in the implementation of the EU's AI Act and its associated code of practice raises significant social concerns, primarily centered around trust and transparency. Trust in artificial intelligence is paramount for both consumers and developers, and the European Commission's intention to build this trust is manifest in its regulatory approach. This delay, however, risks eroding public confidence if the perception that powerful AI technologies are not being adequately regulated takes root. This is underscored by worries that, in the absence of rigorous regulation, AI systems could be open to misuse or misunderstanding by the general public [source](https://www.reuters.com/business/media-telecom/code-practice-help-companies-with-ai-rules-may-come-end-2025-eu-says-2025-07-03/).

Moreover, the delays in implementing the AI Act could hinder the equitable access to the benefits of AI, such as advancements in healthcare and education, which are essential for societal progress. The intention of the AI Act is to maximize these benefits while mitigating associated risks. However, the postponed timeline might decelerate the deployment of such advantageous technologies, leaving society without the enhancements AI promises to deliver in critical areas [source](https://digital-strategy.ec.europa.eu/en/policies/european-approach-artificial-intelligence).

From an ethical standpoint, the delay could extend the timeframe during which AI systems operate with fewer regulations, a situation fraught with potential pitfalls such as algorithmic bias and discrimination. These ethical concerns remain central to the discourse surrounding AI development and deployment, and the European Union's hesitance only prolongs the period in which these issues could persist unchecked [source](https://www.europarl.europa.eu/topics/en/article/20230601STO93804/eu-ai-act-first-regulation-on-artificial-intelligence).

Political Implications and Internal Divisions

The European Union's delay in implementing the Artificial Intelligence Act has sparked significant political implications and exposed the internal divisions among its member states. This delay, driven partly by lobbying from major tech companies such as Google and Meta, highlights the competing interests within the EU framework. These tech giants, along with some European businesses, have pushed for a delay citing the lack of a comprehensive code of practice [Reuters](https://www.reuters.com/business/media-telecom/code-practice-help-companies-with-ai-rules-may-come-end-2025-eu-says-2025-07-03/). This move has not only affected the timeline of the Act but has also caused friction among EU countries, with some member states eager to assert regulatory power while others prioritize economic flexibility and global competitiveness.

At the core of the political implications is the EU's struggle between maintaining its regulatory standards and negotiating the pressures exerted by both large corporations and internal policy disagreements. The decision to delay the AI Act's compliance rules has exposed a rift between those advocating for stringent regulations to ensure ethical AI usage and those who believe that overly restrictive measures could hamper innovation and economic growth within Europe [Reuters](https://www.reuters.com/business/media-telecom/code-practice-help-companies-with-ai-rules-may-come-end-2025-eu-says-2025-07-03/). As some EU countries and tech industries lobby for alterations and postponements, the EU’s ability to present a unified stance on AI regulation comes under scrutiny, potentially diminishing its influence in setting global standards.

Internal divisions further complicate the EU's position in global politics. The lobbying efforts reveal not only the pressures from external actors, particularly the influential tech sector, but also the divergent priorities among EU member states. This discord can weaken the EU's collective bargaining position and give rise to a perception of inconsistency or indecisiveness on the international stage [Reuters](https://www.reuters.com/business/media-telecom/code-practice-help-companies-with-ai-rules-may-come-end-2025-eu-says-2025-07-03/). Correspondingly, this discord may strain transatlantic relations, especially with the United States, where many of these tech corporations are based. Such geopolitical tensions could feed into broader trade discussions, potentially complicating international partnerships.

While the delay offers temporary relief to companies fearful of swift regulatory demands, it emphasizes the underlying political tensions and uncertainties that the EU must navigate. Achieving a consensus on the AI Act's directives has become more complex, as stakeholders within and outside Europe continue to lobby for changes that suit their varied interests. This period of postponement may provide a crucial window for the EU to address and resolve these internal disagreements, ensuring a more cohesive and strategic approach to AI regulation moving forward [Reuters](https://www.reuters.com/business/media-telecom/code-practice-help-companies-with-ai-rules-may-come-end-2025-eu-says-2025-07-03/).

Conclusion

The delay in the implementation of the EU’s AI Act’s code of practice comes at a pivotal moment, highlighting the intricate balance the European Union must maintain between fostering innovation and ensuring effective regulation. This postponement offers temporary relief to businesses, allowing them more time to align with the forthcoming regulations. However, the long-term impact of this delay could be far-reaching. For one, the uncertainty it creates might deter investment and innovation within Europe, as companies may be hesitant to navigate an unfinished regulatory environment [source].

The voluntary nature of the code of practice has stirred debates among industry stakeholders. While it provides legal certainty for companies that choose to comply, those abstaining may face significant legal uncertainties, especially when the AI rules become binding from August 2, 2025 [source]. This dichotomy may pave the way for complex legal challenges as businesses strive to navigate compliance without the guidelines firmly in place.

Moreover, the delay has sparked considerable public and political scrutiny. Criticism has mounted over perceived prioritization of corporate lobbying efforts from tech giants like Google and Meta, thereby potentially undermining the EU's regulatory ambitions [source]. Discontent among experts and the general public focuses on the fear that these delays could eventually benefit larger, especially non-European, tech companies at the expense of smaller European innovators.

The broader implications for the EU in setting the global standard for AI governance cannot be overstated. The delay risks Europe losing its foothold as a leader in trustworthy AI amid an intensely competitive global landscape [source]. Additionally, internal divisions within the EU, highlighted by differing interests among member states and stakeholders, present ongoing challenges in adopting a cohesive regulatory framework that advances both innovation and ethical AI development.

Overall, the path ahead remains fraught with challenges. The EU must navigate the delicate task of adapting to technological advances without stifling innovation, ensuring regulations are robust yet flexible enough to foster growth. The real test will be the EU’s ability to mitigate current delays' possible negative impacts and their potential to erode Europe’s standing in the global AI landscape, while still championing a vision of ethical and fair AI use for all [source].