European AI Act comes into force in the divided tech world

The European AI Act, which went into effect last Sunday, imposes strict rules on artificial intelligence, including a ban on social credit scoring and facial recognition in public places. While Europe is strengthening its regulatory power, the Trump administration in the US is bending back AI legislation. The Trump administration's new term could lead to economic sanctions against European companies: Trump sees EU fines as a tax on US tech giants. Meanwhile, the rise of Chinese companies like DeepSeek continues to increase geopolitical AI tension, especially as they offer technology that challenges US dominance without advanced US chips.

Strict measures and severe fines

With immediate effect, the AI Act prohibits using AI systems to assess citizens based on personality traits such as race or gender. It also bans biometric surveillance and facial recognition in public spaces, with some exceptions for national security. You can find a complete list of all new regulations here. Companies that violate these rules risk fines of up to millions or 7% of their turnover. The legislation affects all organizations operating in the EU, including U.S. tech companies. A second phase of the AI Act will start in August 2025, introducing specific rules for language modeling providers such as ChatGPT.

The impact of AI on global governance: current challenges and future directions

AI technologies influence international policy-making and regulatory frameworks, evolving power dynamics between states and tech companies in a ‘technopolar’ world.

Trump's backlash

Tensions are rising now that Trump has reversed Biden's AI security decree. He considers the EU fines on American tech companies unjustified and labels them a form of tax. They want billions from Google, from Apple, from Facebook. They are American companies, so they shouldn't be allowed to do that, Trump argues. This attitude creates a regulatory gap between the US and Europe, with EU standards for AI security no longer supported by US legislation.

Chinese challenge via DeepSeek

The situation is further complicated by the emergence of DeepSeek, a Chinese AI startup that claims to match OpenAI's capabilities without using advanced U.S. chips. This development has already led to declines in the value of U.S. tech companies. However, DeepSeek's entry into the European market is under pressure - Italy has already blocked the platform due to data transfer concerns, while Belgium and Ireland have launched investigations. The concerns focus on data sovereignty and possible Chinese government access to user data.

International tensions at AI summit

These geopolitical tensions are coming together at the upcoming AI Action Summit in Paris early next week. Representatives from nearly 100 countries, including the U.S. and China, are trying to agree on the safe development of AI. Philanthropic institutions and companies are expected to pledge huge sums for AI projects. The summit represents a crucial test of international cooperation on AI regulation.

Europe's unique position

The current situation also presents Europe with opportunities. With strict regulation and expertise in regulated industries, the EU can position itself as a leader in trusted AI development. For many European tech companies, DeepSeek also offers an opportunity to close the gap with the U.S. market. While nearly $100 billion was invested in AI companies in the U.S. by 2024, it was only $15.8 billion in Europe. DeepSeek now offers an alternative for companies that generally cannot afford much in the way of AI development.

These European startups benefit from sharp cost advantages DeepSeek. But there are also concerns

European companies benefit from sharp cost advantages, but ethical concerns and security issues remain a challenge.