On August 2, the new European Union law on artificial intelligence will come into force, with rigorous controls on content generated by this technology, its identification at all times, as well as fines for non-compliance.

Monitoring of all AI-generated content is planned, which must be clearly labeled with visible watermarks, whether image, audio, or video.

Furthermore, the new law will require mandatory transparency; this means that “users must know at all times if they are interacting with a machine (chatbots) and that deepfakes must be identified.”

There are limited exceptions for the technical markup or visual labeling of “texts whose use is evident from the context and journalistic pieces that have undergone human review and editorial responsibility.”

Regarding penalties, the law establishes fines for non-compliance with transparency regulations that can reach 15 million euros or 3 % of the annual global revenue of the company involved.

The affected companies in the sector include OpenAI with ChatGPT, Anthropic with Claude, and Google with Gemini; also, image generators like Midjourney, speech synthesis tools, and other systems that allow the use of AI for creation.

M.Pino

Source: elpais

(Reference image source: Levart_Photographer on Unsplash)

Follow our news on Google! For current, interesting, and accurate information, click here to see all the content on Bitfinance.news. You can also find us on X/Twitter and Instagram

La entrada EU AI Act to come into force on August 2 apareció primero en Bitfinance.

This Thursday, August 1, the AI ​​Law came into force in the European Union. There will be an adaptation period for companies until August 2026 and fines of up to 7% of annual turnover are expected for non-compliance.

The law, which was announced last June, came into effect this Thursday, establishing regulations for everything related to AI in the region. It is stipulated that in a period of six months, “starting on February 2, 2025, the general provisions and prohibitions of unacceptable risk must be complied with.”

On the other hand, in nine months, starting on May 2 of next year, the codes of good practices must be complied with; and in 12 months, starting August 2, 2025, “the general rules must be complied with, and the different countries must have updated their rules regarding sanctions and fines.”

In accordance with these guidelines, from 2025 the legislation would be ready to apply the corresponding fines to companies that do not comply with these regulations.

It is important to clarify that a longer time frame will be given for high-risk systems, the concept referring to “AI systems that may have a considerable detrimental effect on the health, safety and fundamental rights of people.” In this case, a period of 36 months will be granted to comply with the law. So it will be from August 2027 when “these AI models will be forced to document all of their operation and be totally transparent about how they work.”

Regarding fines, there is talk of 7 % of the annual turnover with the intervention of the European Office of Artificial Intelligence, in charge of enforcing the law, “preparing reports, updating requirements and providing support to companies to be able to adapt correctly.”

M.Pino

Source: xataka

(Reference image source: Unsplash+, in collaboration with Getty Images)

Visit our news channel on Google News and follow us to get accurate, interesting information and stay up to date with everything. You can also see our daily content on Twitter and Instagram

La entrada This August 1 the AI ​​law came into force in Europe apareció primero en Bitfinance.