A large organization composed of big technology companies and some business giants warned the European Union on Thursday local time not to over-regulate important systems in the field of artificial intelligence known as basic models in the upcoming artificial intelligence rules, because this may stifle emerging start-ups or force some AI technology leaders to leave the EU. The call comes as EU member states and EU lawmakers enter the final stages of negotiations on rules that could set a baseline for how other countries regulate models underlying artificial intelligence.

The biggest focus of debate between enterprises and technology groups and the EU is the basic AI model. For example, the GPT-4 large language model (LLM) behind OpenAI's ChatGPT is a basic AI model. similarThe basic model of GPT-4 is an artificial intelligence system that has been trained thousands of times based on large amounts of data and has the powerful ability to learn from new data to perform various tasks.

For example, the GPT-4 large language model uses a large amount of data for pre-training, and users can use ChatGPT to generate text, answer questions, translate languages, write articles and other language-related tasks, so it plays an important basic role in the AI ​​field. GPT-4 has a larger model size and stronger performance than its predecessor GPT-3, and can more accurately understand and generate complex language content.

It is understood that the main members of the "Digital Europe" formed by large technology companies and business giants include Airbus, Apple, Ericsson, Google, the London Stock Exchange, SAP, etc. The joint organization said in a letter: "In order for most European countries to eventually become global digital industry powers, we urgently need unicorn companies that can use basic models and GPAI to lead the wave of artificial intelligence innovation."

“As diverse representatives of Europe’s digital industry, we see huge development opportunities arising from the foundational models of AI, as well as new innovative players emerging in this space, many of whom were born and raised in Europe. Let’s not impose strong regulations on them before they have a chance to scale, and let’s not force them to leave.”

It is understood that up to 32 related digital industry associations covering all Europe have also signed this letter. GPAI refers to general artificial intelligence system.

Participants who signed the letter stated that only 3% of the world's artificial intelligence unicorn companies are from EU member states, which shows that the EU's financial support plan for AI start-ups in Europe and around the world is crucial. They support a joint proposal by France, Germany and Italy to limit the scope of artificial intelligence rules for basic models to transparency requirements.

They also said that the scope of the current draft artificial intelligence rules is very broad and may conflict with existing legislation and regulations in some traditional fields such as health care.

"We are increasingly frustrated by the lack of interest we see in the relevant impact in the healthcare industry. Our impression is that people no longer care about the content, they just want to get things done quickly. And we are just collateral damage in these tasks." Siemens Healthineers spokesperson Georgina Prodhan said.

The large group of companies has also resisted calls from creative industries for AI rules to address copyright issues. “The EU’s comprehensive copyright protection and enforcement framework already contains specific provisions to help address copyright issues related to artificial intelligence, such as text and data mining exemptions and corresponding provisions,” they said.

Related articles:

The EU is mired in internal strife over AI regulation. The German-Italian-French agreement was denounced by European lawmakers as a declaration of war.