A group of businesses and tech companies have issued a joint letter to European Union regulators warning against over-policing powerful artificial intelligence (AI) systems at the expense of innovation.
The letter, which was sent on Nov. 23 and undersigned by 33 companies working in the EU, stresses that too-stringent regulations for foundation models like OpenAI’s ChatGPT and general-purpose AI (GPAI) could drive necessary innovation from the region.
It highlighted data that shows only 8% of companies in Europe use AI, which doesn’t come close to the European Commission’s 2030 goal of 75%. Additionally, only 3% of the world’s AI unicorns come from the EU.
“Europe’s competitiveness and financial stability highly depend on the ability of European companies and citizens to deploy AI in key areas like green tech, health, manufacturing or energy.”
The companies stressed that for Europe to develop into a “global digital powerhouse,” it needs companies leading in AI via foundation models and GPAI — two AI technologies under scrutiny in the forthcoming EU legislation.
“Let’s not regulate them out of existence before they get a chance to scale, or force them to leave.”
Related: Greece establishes AI advisory committee to create national strategy
In addition to stressing the importance of not over-regulating the technologies, the companies also suggested solutions for EU leaders.
Suggestions include reducing compliance costs for companies, focusing on regulating high-risk use cases and not specific technologies, and clarifying where there are already overlaps in existing legislation.
This development comes as the EU is working on finalizing its landmark EU AI Act, which was initially passed back in June and is currently undergoing review and revision by member states.
Shortly after the initial act was passed, another letter was signed by 160 executives in the tech industry warning EU officials about the implications of too-strict AI regulations.
Magazine: ‘AI has killed the industry’: EasyTranslate boss on adapting to change
