The EU will issue more restrictive rules for large AI models: report

European Union representatives are negotiating plans to introduce additional regulations on large-scale artificial intelligence (AI) systems, according to a Bloomberg report.

The European Commission, the European Parliament and various EU member states are said to be discussing the potential impact of large-scale language models (LMS), including Meta Llama 2 and OpenAI's ChatGPT-4, and possible further restrictions. As part of the upcoming AI Act.

According to Bloomberg, sources close to the matter said the goal is not to burden new startups with more regulations while waiting for larger models.

According to the sources, the agreement reached by the negotiators on the topic is still at an early stage.

The AI ​​Act and the new proposed regulations for LLMs will have a similar approach to the EU's Digital Services Act (DSA).

EU lawmakers recently implemented the DSA, which requires platforms and websites to have standards for protecting user data and scanning for illegal activity. However, the web's largest forums are tightly controlled.

Companies under this category, such as Alphabet and Meta, had until August 28 to update their services to comply with the new EU requirements.

Related: UNESCO and the Netherlands design an AI control project for the EU

The EU's AI legislation is one of the first mandatory laws set by a Western government. China in 2010 It has issued its own AI rules, which came into force in August 2023.

Under EU AI regulations, companies that develop and deploy AI systems are completely prohibited from conducting risk assessments, labeling AI-generated content, and using biometric tracking.

However, the law has not yet been passed, and member states still have the ability to veto any of the proposals presented by the parliament.

In China, since implementing its AI laws, more than 70 new AI models have reportedly been released.

Magazine: The Truth Behind Cuba's Bitcoin Revolution: An On-the-Ground Report