EU backslides on AI rules – what does it mean for Switzerland? 

The EU wants to relax its rules on privacy and AI. For many critical voices, this is a concession to Big Tech and the Trump administration. 2025 Thierry Monasse

The European Commission wants to delay implementing parts of its new Artificial Intelligence Act, giving big tech companies more time to harness citizens’ data. The move could also affect Switzerland, whose digital policies often follow Brussels’ lead.  

In the midst of uncertainty that reigns in the world of artificial intelligence (AI) there was, until recently, one certainty: the European Union was one of the rare institutions ready to challenge Big Tech companies with strict laws to protect its citizens. 

This certainty crumbled when the EU Commission announced a planExternal link on November 19 to simplify some of its digital regulations. Among the most important are two that regulate privacy rights and artificial intelligence tools: the General Data Protection Regulation (GDPR) and the Artificial Intelligence Act (AI Act). 

On paper, the plan, called “Digital Omnibus”, aims to help European companies “do business” by easing paperwork and unlocking access to “high-quality data” to enable AI training, as stated in the official press releaseExternal link. 

The move aims to allow companies to comply with complex European regulations without penalising innovation. But the legal changes proposed by this plan are worrying because they weaken the privacy and security guarantees of AI systems, according to several analysts. Austrian lawyer and activist Maximilian Schrems calledExternal link the EU operation “the biggest attack on the digital rights of European citizens in recent years”. 

Although it is not part of the EU, Switzerland is partially integrated in the European single market and the EU represents its most important trading partner. The decisions taken in Brussels can also indirectly influence the Swiss population, as well as its legislators. 

The EU caved to US pressure 

The proposal to delay the implementation of rules concerning high-risk AI, in some cases until 2027, is especially controversial. These systems, such as those used for biometric identification and screening of job applications, can endanger the health, safety and fundamental rights of citizens. Many observers view this as the first step in an effort to dismantle a law that mainly targets large US companies and that the Trump administration has spoken out against. 

“No one is fooled by the transatlantic origin of these attempts,” saidExternal link Thierry Breton, former European commissioner for the internal market and services and one of the major contributors to the AI Act. 

The loosening of the EU’s legislative fabric comes as Brussels is engaged in talks with the US administration over reducing tariffs on key products for the bloc’s economy, including wine, spirits, steel and aluminum. Howard Lutnick, the US secretary of commerce, has already publicly statedExternal link that a beneficial agreement in this direction is conditional on a relaxation of technological regulations. Given the timing, it seems plausible that the EU has bowed to US pressure. 

“The US protects their tech industry, which they consider to be one of the most critical infrastructures,” David Vasella, a Swiss attorney and expert in data privacy and technology law, told Swissinfo. For Vasella, the fact that the EU may be losing ground on digital regulations is not a good sign for its citizens overall. It suggests that the bloc is becoming weaker economically and more dependent on Washington, he says. 

Companies rejoice 

At the same time, the EU has also heeded appeals from dozens of European companies concerned about the difficulty of adapting to the new AI rules. Companies such as Airbus and Lufthansa, which also owns Swiss International Airlines, asked in an open letterExternal link that the EU suspend implementation of the law for two years. This was to give those doing  business in Europe breathing space to adapt to the EU’s “increasingly complex laws”. 

A similar warning had already come in 2024 from former European Central Bank President Mario Draghi. In his reportExternal link to the EU Commission, Draghi had pointed out that excessive regulatory complexity was hampering European competitiveness and innovation. 

Swiss companies and experts have also warned  about the significant costs of complying with EU rules and continuing to sell products and services in the single European market. For this reason, leading figures in Swiss AI such as Marcel Salathé, co-director of the AI Center at the Swiss Federal Institute of Technology in Lausanne (EPFL), welcomed the simplified European legislative plan. 

“I am happy to see that the EU is reconsidering its data and AI regulations,” Salathé wrote on LinkedInExternal link. He called the GDPR “overly burdensome” and the AI Act “a premature sea of vagueness”. 

Vasella tends to agree with this view and thinks that some simplifications are reasonable and justified. “Fixing some very burdensome provisions makes a lot of sense,” he says.   

Concerns also for Swiss citizens

While simplifying regulations makes companies happy, citizens have less reason to celebrate. One of the most worrying changes in the Digital Omnibus concerns the facilitation of the use of personal data to train AI models. This means that in many cases companies would be able to exploit such data without having to obtain explicit consent from users. The reform would also revise the definition of ‘personal data’, making it possible for companies to use anonymised data as long as individuals cannot be identified. 

Even if European laws have no direct impact on Swiss citizens, it is hard to imagine that physical borders can protect the Swiss population from data abuse in the digital space. Especially when we consider the numerous online services provided by companies located in the EU. According to Vasella, delaying the regulation of high-risk AI systems is dangerous even for Swiss citizens, because harmful applications will continue to develop rapidly.  

Bern caught in the crossfire  

The last few years have also shown that Swiss legislators are very sensitive to developments beyond its borders, particularly in Brussels. The GDPR prompted Swiss politicians to revise the Federal Data Protection Act, strengthening guarantees in the event of a privacy breach. And the AI Act has forced the Swiss government to consider more precise regulation of artificial intelligence in the face of risks such as workplace discrimination and unequal access to essential services.  

Switzerland is therefore exposed to two trends: the relaxation of European rules and the growing influence of the United States. The government has already signalled its intention to yield to Washington, by promising in a Joint StatementExternal link not to introduce digital services taxes and to ease data transfers to the US as part of customs negotiations.  

In this complex geopolitical landscape, Big Tech companies are the ones that gain the most, as they continue to shape citizens’ daily lives mostly unchecked. “The real threat to citizens in the EU and in Switzerland are purposefully addictive apps like TikTok. We let them flood our markets,” Vasella argues. 

Faced with the backtracking of its influential neighbour and growing pressure from across the Atlantic, Bern could find convenient pretexts to further slow down – if not quietly stall – the process of adapting its laws. The Swiss “wait and watch” approach has never been more convenient. 

Edited by Veronica De Vore/ac