European Industry Pushes Back on the EU AI Act – Key Takeaways for Employers

Over 45 leading European companies just publicly urged the EU to pause the rollout of the EU AI Act’s impending rules for high-risk and general-purpose AI (GPAI) systems, citing concerns over regulatory complexity and competitiveness. A July 3 letter from the consortium comes less than a month before tools like GPT-4, DALL-E, Google BERT, and Midjourney 5.1 will become subject to new rules, potentially disrupting innovation and AI ambitions of many European countries. While the European Commission has emphatically rejected attempts to delay implementation of the August 2 effective date, it has also indicated it might pause the implementation of some parts of the AI Act if standards and guidelines to implement the law are not ready in time. What impact will this letter have, and what should employers doing business in Europe do?

A Quick Refresher on the EU AI Act

Here’s a quick primer about the EU AI Act. Passed in March 2024, it’s the world’s first comprehensive framework regulating artificial intelligence – and it has significant implications for employers operating in the EU.

It classifies AI systems by risk and imposes strict restrictions on high-risk applications (which include most employment-related activities like recruitment, job advertising, screening, hiring, HR, and worker management), while requiring transparency and accountability for others. Its requirements unfold in stages through 2026, but some critical new rules will come online starting August 2. Specifically, employers that modify common GPAI models risk being classified as “providers” and could face full compliance obligations, including being required to audit all AI systems used in HR and management.

Companies operating in or serving the EU must comply or risk fines of up to 7% of global revenue. You can read our full summary of the EU AI Act here.

The EU Industry’s Call for a Pause

Last week, more than 45 of Europe’s most influential companies – including Airbus, ASML, Lufthansa, Mercedes-Benz, Siemens Energy, and Mistral – called on the European Commission to “stop the clock” for at least two years on the AI Act’s most stringent requirements. Their open letter targets rules impacting both high-risk AI systems (effective August 2026) and GPAI models (rules begin August 2025), warning that the current approach threatens Europe’s competitiveness and innovation.

• Implementation Uncertainty: With key standards and guidelines still in development, companies fear they cannot comply effectively.
• Regulatory Overload: Companies argue that the AI Act’s “unclear, overlapping and increasingly complex” rules disrupt the EU’s traditional balance between regulation and innovation.
• Competitiveness at Risk: The signatories warn that the current regulatory approach could undermine Europe’s global standing, affecting not just tech but all AI-dependent industries.
• Request: The companies seek a two-year delay on the rules governing high-risk AI systems and GPAI so regulators can develop relevant guidelines and allow industries to find viable solutions.

The EU’s Response: No Blanket Delay, But Flexibility Possible

Despite industry pressure, the European Commission has so far rejected the idea of a broad pause. Implementation deadlines for GPAI (August 2) and high-risk AI systems (August 2026) still remain in place.

However, the Commission’s digital chief, Executive Vice President Henna Virkkunen, recently signaled that targeted delays could be considered if essential standards and guidelines are not ready on time. This suggests a potential, though limited, willingness to adjust timelines for specific provisions if practical hurdles persist.

Quick Breakdown: High-Risk AI and GPAI Models

High-Risk AI Systems AI systems used in employment—such as recruitment, job advertising, application screening, performance monitoring, and worker management—are classified as high-risk under the AI Act. You can read more about the AI classification systems and their effects on HR, hiring and employee management here. Obligations Beginning in August 2026 Register high-risk AI systems in an EU database Conduct fundamental rights impact assessments Complete conformity assessments and maintain technical documentation Implement robust data governance and bias mitigation Ensure transparency and human oversight Maintain quality management and cybersecurity standards Practical Impact Employers must audit all AI systems used in HR and management to determine if they fall under the high-risk category. Non-compliance can result in fines up to 7% of global revenue.

 

General-Purpose AI (GPAI) Models GPAI models – like GPT-4, DALL-E, Google BERT, and Midjourney 5.1 – are foundational AI systems that can be adapted for a wide range of applications, including workplace tools. Key Dates August 2, 2025: Obligations for providers of new GPAI models begin August 2027: Obligations for operators of GPAI models already on the market take full effect Employer Risks Employers using third-party GPAI are considered “operators” and face lighter obligations. If an employer develops or modifies a GPAI model (e.g., by fine-tuning with new data), they may be reclassified as a “provider,” triggering full compliance requirements.

What Employers Should Do Now

1. Audit AI Systems

Identify all AI tools used in HR, management, and operations. Determine if any are high-risk or involve GPAI models.

2. Be Cautious with Fine-Tuning GPAI Internally

Keep in mind “fine-tuning” or significantly modifying GPAI models in-house could make your company an AI provider, which would trigger significant additional obligations. Let external providers handle modifications if possible to avoid being classified as a provider under the Act. If you must modify in-house, be ready to comply with heightened obligations.

3. Review Vendor Agreements

Ensure third-party AI vendors are compliant with EU regulations. Do not rely solely on vendor assurances – conduct your own due diligence.

4. Prepare for Compliance

Begin preparing documentation, impact assessments, and governance structures now, as guidance evolves.

5. Monitor Regulatory Updates

Stay informed about new guidance, codes of practice, and enforcement actions as the regulatory landscape continues to shift.

Conclusion

We will continue to provide the most up-to-date information on AI-related developments, so make sure you are subscribed to Fisher Phillips’ Insight System. If you have questions, contact your Fisher Phillips attorney, the authors of this Insight, or any attorney in our AI, Data, and Analytics Practice Group or International Practice Group.