Member countries of the European Union and legislators from the European Parliament have struck a deal on a diluted version of major artificial intelligence (AI) regulations within the European Commission's digital omnibus package.
The preliminary deal involves postponing certain AI system obligations to avoid legal ambiguity, the European Parliament stated in a press release early Thursday following late-night discussions.
A consensus was established to delay regulations on high-risk AI systems, such as those pertaining to biometrics or linked to critical infrastructure, education, employment, law enforcement, and border management, to December 2, 2027, changing the previous deadline of August 2 of this year.
The omission of machinery from the AI act was also agreed upon, based on the belief that it only needs to adhere to existing sector-specific safety regulations.
Also Read: Google Proposes Spam Policy Changes to Avoid EU Fine
Critics view the action taken by EU co-legislators as succumbing to demands from corporations.
A compulsory watermark for AI-generated content will be enforced starting December 2.
Also Read: EU and Capitals Deadlock Over Proposed US Deal
"The agreement reached today regarding the AI act greatly aids our businesses by lowering ongoing administrative expenses." "It guarantees legal clarity and facilitates a more seamless and uniform application of the regulations throughout the Union, enhancing the EU's digital autonomy and overall competitiveness," states Marilena Raouna, Deputy Minister for European Affairs of Cyprus.
EU member states and negotiators from the European Parliament also reached an agreement to ban AI applications utilized for generating unauthorized sexually explicit deepfakes.
The material might consist of pictures, video, or sound. Firms have until December 2 of this year to align their systems.
Also Read: IREN Expands AI Cloud Platform to Europe
The proposed prohibition is specifically aimed at including the production of content that illustrates child sexual abuse.
Through this action, the EU is addressing several factors, including events from earlier this year, when individuals exploited US billionaire Elon Musk's chatbot Grok to create and spread millions of AI-generated sexually explicit images of women and children across the internet.