The EU reportedly plans to introduce a novel three-tiered system to regulate artificial intelligence systems and models, potentially making it the first Western government to do so.
What Happened: The EU’s proposed regulatory scheme would classify and control AI technology according to its power and adaptability. The system would require different levels of external testing for various AI systems, making the EU the first Western government to enforce mandatory rules on AI, reported Bloomberg, citing the proposal document.
The AI Act, the proposed legislation, would necessitate AI systems for predicting crime or sorting job applications to undergo risk assessments, among other prerequisites. The goal is to finalize the legislation by year-end, following a meeting scheduled for Oct. 25.
See Also: This ChatGPT Jailbreak Lets You Get AI’s Help To Roast People
The proposed system comprises a first category for all foundational AI models.
A second tier for “very capable” systems would differentiate their abilities based on the computing power employed to train them.
The final category — general-purpose AI systems at scale — would encompass the most widely used AI tools and be gauged by the total number of users.
While the proposed regulations are still open to change as discussions continue, the report noted that representatives from the EU’s three institutions generally supported a tiered approach at a meeting earlier this month.
Why It’s Important: This proposal comes as EU officials are contemplating stricter regulations for large AI system operators like ChatGPT-parent OpenAI and Mark Zuckerberg-led Meta Platforms under the upcoming AI Act.
The potential rules for large language models could impact major AI tools by these giants. However, the agreement is still in the early stages, and details are being discussed.
While the EU has been pushing for more AI regulations, the U.S. is also participating in international discussions. Vice President Kamala Harris is set to represent the U.S. at a U.K. summit focusing on AI safety measures next month.
Since OpenAI launched ChatGPT last year, AI tools that can generate texts, images, and videos have become incredibly popular. This has made many tech companies rush to make their own versions of these AI tools.
However, users’ feelings about AI have changed a lot.
At first, many were excited about it, but now, some are worried because of the fast-paced speed of AI advancements.
Image Via Shutterstock
Check out more of Benzinga’s Consumer Tech coverage by following this link.
© 2024 Benzinga.com. Benzinga does not provide investment advice. All rights reserved.
Comments
Trade confidently with insights and alerts from analyst ratings, free reports and breaking news that affects the stocks you care about.
