Open letter: For uniform AI regulation across the EU

An open letter signed by a number of companies and researchers working in the EU and developing and providing AI services says the EU is at risk of missing the age of artificial intelligence. The question is what data can be used to train AI models. The signatories warn that if no data from the EU flows into the most widely used models in the world, the models will also lack an understanding of the EU’s people, languages ​​and cultures.

Advertisement

The letter appears to have been well-advanced by Meta. Mark Zuckerberg himself is one of the signatories. The letter said, among other things:: “The EU will also miss out on other innovations, such as Meta’s AI assistant, which is on its way to becoming the most used AI assistant in the world by the end of this year.” Meta’s AI models are open source, and Mark Zuckerberg has already said several times that he considers this path to be the only viable option in the field of AI. According to Zuckerberg, an entire ecosystem will be able to develop around the open source model.

Signatories include SAP, Spotify, Thyssenkrupp, EssilorLuxotica, Artefact and numerous scientists as well as civil society and trade associations.

Criticism of GDPR, not the AI ​​Act

However, what is being criticised is not, as one might initially think, the AI ​​Act, i.e. the AI ​​regulatory framework that has recently come into force. The open letter is primarily about data protection. The Irish Data Protection Authority (DPC), which, among other things, is responsible for Meta in the EU, had banned the company from using people’s data on Facebook and Instagram for AI training. Meta wanted to be reassured by opting out that they could access all posts, but not private messages. Although the DPC initially allowed an opt-out process, it reversed this decision after protests. Meta stopped using the data.

Cybersecurity Act: rebuke from the Federal Audit Office | Heise Online

In fact, it can be assumed that all other providers of large AI models already use public contributions on platforms to train their models – without consent. OpenAI’s CTO Mira Murati confirmed in an interview that they used everything publicly available. Whether this is legal or not is yet to be decided. This applies to social media posts as well as YouTube videos, newspaper articles and works of art. The latter is about copyright and not about data protection.

According to the open letter, the EU risks losing connectivity and weakening economic growth. “Research shows that generic AI could increase global GDP by ten percent over the next decade, and this growth should not be withheld from EU citizens.”

To avoid this, the signatories call for “harmonized, consistent, prompt and clear decisions within the framework of EU data protection rules that enable European data to be used in AI training for the benefit of Europeans.”

(EMW)