AI-Reporter: ‘European Commission lets its ears hang to Techoligarchs and Trump’

It is one of the hottest European files at the moment: the European AI Act. The package with laws that must curb artificial intelligence in the coming years is not yet two months in force, and part of it is already under discussion. That was apparent on Tuesday during a meeting where legislators, rapporteurs and the European AI office came together to talk about the elaboration of the measures.

The tone was ‘rather heated’, Kim van Sparrentak, MEPs, agrees for GroenLinks-PvdA and AI reporter from the first hour. The meeting followed a « consciously timed » fire letter addressed to the European Commission, signed by the negotiators of the AI ​​Act in the European Parliament. They express ‘great concern’ about the latest version of the code of conduct in the new AI Act.

Especially the relaxation of the risk assessment requirements for the largest tech companies is difficult for MEPs. The obligation for the companies behind the largest AI systems-think of chatgpt from the American OpenAi and Copilot (Microsoft) or the Chinese Deepseek-to test their content on discriminatory or racist expressions. « That was never the intention of the law, » said the co -legislators.

« The largest AI providers do not oblige to estimate and limit system risks that lead to discrimination, is dangerous, undemocratic and creates legal uncertainty to fully reinterpret and limit a legal text that has been reached, » said the letter, which is partly signed by Van Sparrentak.

The European commsion has been increased from the tech lobby?

« Yes, she lets her ears hang to Techoligarchs. She rolls out the red carpet for foreign influence of elections via AI and allows discriminatory systems to the European market. We totally supply the EU to the whims of tech bosses.

European Commission is rolling out the red carpet for foreign influencing of elections via AI

„De techlobby zit hier al vanaf het begin bovenop. Ze zijn altijd tegen de regulering van kunstmatige intelligentie geweest. ‘Wij gaan überhaupt niet tekenen’, heeft onder andere Meta gezegd. Dreigingen van grote techbedrijven dat ze niet mee willen doen aan de AI-wet, zijn aan de orde van de dag. We horen nu dat de Commissie, om maar geen gedoe te krijgen met de techlobby en de VS, de wet heel nauw wil interpreteren zodat That does not turn out too strictly. « 

What’s at stake?

« Our fundamental rights. As a shadow ripor, I have made a neck case of adding rules that prevent discrimination. The Netherlands knows better than anyone how these types of systems can discriminate, just think of the allowance affair. The committee now wants to agree to legislation that does not take such examples into account. On Tuesday it was not to be forbidden about humanity. Hollywood-like, Terminator scenes, but about fundamental human rights that are at stake. ”

Is this a gesture to President Donald Trump, as part of a larger negotiating game?

“Absolutely, the tone of the committee has changed after the AI ​​summit in Paris in February. Commission chairman Ursula von der Leyen spoke with the American vice president JD Vance and it will not surprise me that the AI ​​legislation was deployment of the conversation, as part of larger geopolitical forces. The White House is currently ignored and the White House is now ethical and the White House is now ignored. After five years it works and whether something has to be adjusted, but Brussels is now in the spell of Deregulation. « 

The fire letter comes out when the Commission is tapping into companies such as Google and Apple, to better comply with the European Digital Market Act DMA. Does Brussels not issue a double signal?

« In that issue it is more about abuse of power by the large companies, although the AI ​​companies are also working hard to create a monopoly. Yet it is slightly different for them. But we also see the pressure from the US, and more and more from the right-wing European Parliament, to da. Being overjoyed with the law, because they increase their chances to end up in the app stores. « 

What steps want and can you still take?

« If we want to do everything we can to protect our democracy within Europe and to guarantee human rights, we should not bow to the tech lobby. If you bring a car or new medicine to the market, they must also be safe. The products of these tech companies must be treated just like that of other companies.

« The law states that large AI models should be tested for, for example, reliability and risks for manipulation during elections. But also on matters that can pose a danger to safety and health, such as the spread of content about eating disorders and child abuse. Tech companies must prove that to prevent this obligation. Pessors of the proposal is now on May 2, but the question is whether it will be achieved. « 

To share

Mail the editors