“The AI industry is very emerging and we would probably be able to make better choices if we were waiting for five years,” he said.
Another battlefield is a voluntary set of rules for providers of AI models for general use, which are the most complex AI models such as the GPT of Openai and the Gemini of Google, where AI companies have sounded the alarm bell on the rules transforming into another burden.
But civil society groups repel deregulation, and this is likely to make discussions on changes in AI laws as heavy as the first time.
The language of Wednesday’s strategy was debated until the publication: the part on the simplification of the final version was strongly softened in relation to a disclosed project reported by Politico before the announcement. A line that said there was an “opportunity to minimize the potential burden of conformity” was replaced by a declaration that there is a “need to facilitate compliance with the AI Act”.
Two main EU legislators also postponed Wednesday against the previous decision of the Commission to chop a proposal for a single EU responsibility system for damage caused by AI. They criticized the reasoning of the commission as “premature and unconvincing”.
Maximilian Gahntz, leader in AI policy in Mozilla, warned that a push for the rules of the simple AI “should not lead to undermining the effectiveness of EU AI of AI and what they had to accomplish.”
“Simplification should not mean deregulation,” he said.
Politices
