ASML, Mistral Ask EU to Delay Start of AI Act Rules
(Bloomberg) – European companies, including Asml Holding NV and Mistral AI, called the European Union to suspend the implementation of the arrangement of artificial intelligence because they said that the continent was at risk.
In a letter signed by representatives from more than 45 organizations, the President of the European Commission, the President of the European Commission, asked Ursula von Der Luyen to postpone the implementation of the rules in the most powerful AI models and called for a “innovation -friendly approach”.
“When this postponement is combined with the commitment to accelerate regulatory quality, it will send a strong signal to innovators and investors in the world that it is serious about the simplification and competitiveness of Europe,” he said.
Other enterprises that signed the letter include Airbus SE, Mercedes-Benz Group AG, BNP Paribas, Deutsche Lufthansa, Publicis Groupe SA, Prosus NV and Siemens Energy Ag. They called for a delay for two sets of rules applied to General Purpose AI models and high -risk AI systems.
Businesses complain that the commission cannot provide basic instructions and standards, including an application code that provides instructions to comply with the arrangement of advanced companies. Preparation of the Rules of Application – It is already late after the commission was planned for May. Technology companies have accused this of going beyond the EU’s AI Law and creating its own troublesome rules. The US government was also weighed and sent a letter to a series of European countries that called for the commission and the block code to scrapping in April.
Now, these delays threaten to completely reclaim the implementation of the law. According to the EU’s gradual deadline for the law, the most powerful AI models of Openai, such as chatgpt, should be suitable until August. However, as of July, the Commission’s AI developers, academics and work groups of digital rights activists were still in a detailed guidance.
The AI law adopted last year is a series of controls to prevent the worst abuse of technology. AI developers will need to provide information about how their models are trained and their policies that respect the copyright law. The most sophisticated models should take additional steps to reduce risk, increase security and give information about system architectures. The law also puts railings on some types of use, such as restrictions on tools that define people in public spaces in real -time.
Although the application code is voluntary, it provides a framework that will help companies to remain in accordance with the Law. Chewing these rules may be fined 7% of a company’s annual sales or 3% for companies that develop advanced AI models.
Speaking almost at a Brussels event in February, Meta Platform Inc.’s Global Affairs Chairman Joel Kaplan described the rules of application as “useless and unmissable ve and added that the company would not sign it in its current form. Alphabet Inc. He also criticized the rule of the rule, and a manager said in an interview with Politico in February that he took the requirements of third -party model tests and copyright protection very far.
(Corrects the removal of SAP from the list of signs, changes the number of people who sign)
There are more stories like this Bloomberg.com
