EU Agrees Landmark AI Regulation Deal

The European Union (EU) has reached an unprecedented agreement on global-level legislation to regulate artificial intelligence (AI) after three days of intense negotiations between member states and the European Parliament, according to Commissioner Thierry Breton.

French Commissioner in charge of digital affairs, Breton, made the announcement via the Ex platform, stating that a political agreement has been reached regarding legislation aimed at fostering innovation in Europe in advanced technologies while mitigating potential misuse.

Breton added that with the political agreement on the AI law today, the EU becomes the first continent to establish clear rules for the use of artificial intelligence.

According to a report by Agence France-Presse, these regulations, closely monitored globally, will not only impact major AI system developers like "Meta," "Google," "Microsoft," and "OpenAI," but also the technologies that these systems may be used for in educational, healthcare, banking, and even judicial and legal sectors.

The new legislation includes rules to ensure the quality of data used in algorithm development, verifying that it does not violate copyright law.

It will also define standards for harnessing the potential benefits of this technology while protecting against risks such as job automation, the dissemination of misleading information online, and threats to national security.

While the law still needs to go through a few final steps for formal approval, the political agreement signifies that the broad outlines have been defined.

European policymakers have focused on the most critical uses of AI by companies and governments, including those used in law enforcement or operating critical services in the energy and water sectors.

The legislation will require generative AI systems like "ChatGPT" or those that create manipulated images to provide greater transparency, disclosing to people that the final content was generated by artificial intelligence, according to the legislative draft.

The new legislation will also impose restrictions on facial recognition programs by law enforcement and governments, with exceptions tied to national security, and companies providing such software that violate the new regulations may face fines of up to 7% of their global revenues, the newspaper reported.

While the new law is welcomed as an organizational breakthrough, questions remain about its effectiveness, as some regulations will not take effect until 12 or 24 months, a long timeframe for the ever-evolving field of artificial intelligence.

Implementation mechanisms also remain unclear as the law covers 27 countries and requires the appointment of new experts at a time when budgets are limited.