2024-10-31
After what can only be described as a legislative marathon, the European Union passed its Artificial Intelligence Act, or “AI Act”. It is the world’s first attempt to set legal guidelines on how humans should regulate AI. But how will it provide oversight over AI models like ChatGPT? And despite all the EU’s efforts in pushing the law through, why is Europe less than thrilled about its efficacy?
In 2022, ChatGPT, the first of many modern iterations of generative AI, came into existence. The EU responded by drafting its Artificial Intelligence Act, or “AI Act”. After many revisions, the law has finally come into effect.
According to the official website of the Council of the European Union, the AI Act is a world-first milestone aimed at providing oversight on AI.
The EU's AI Act classifies AI systems by evaluating their risk of causing harm.
In August, the AI Act came into effect. Its influence has caused ripples across multiple fields of AI applications.
The law came from good intentions and is a global trailblazer. But it has raised serious concerns in Europe.
For one thing, the AI Act only addresses the basics of what's at stake—and even that took nearly two years of negotiations—but it lacks details on the specifics of actionable regulatory measures.
For example, according to the new law, AI applications such as facial recognition need to be tested before deployment. Certain parameters must be adjusted to see if the programs will still work. But the AI Act failed to provide clear guidelines on how these tests should be conducted.p>
"The devil will be in the details," opines a diplomat who took a leading role in drafting the bill. "But our legislators were tired, and the timeline was tight."
Others worry that oppressive regulations will stifle new AI innovation.
Andreas Cleve, CEO of the Danish medical startup Corti, has a lot of ideas for AI-based entrepreneurship. For starters, he wants to convince clinical doctors to use the "AI copilot" developed by his company. But he worries that the new law will dash his hopes, since the compliance fees the AI Act requires of his company of 50 workers could amount to six-figure sums.
"I worry about legislation that becomes hard to bear for a small company that can't afford it," sighs Cleve. "It's already a daunting task to raise cash. Now you have this tax imposed. We also need time to understand it."
The Financial Times does not mince words. "Entrepreneurs are doubtful about the EU's ability to become a superpower in AI while implementing the new rules," it reports.
But it is precisely because of this that some critics worry that the new regulations are skewed in favor of business leaders and overlook certain risks.
A high-ranking official in the EU states that the enterprises lobbying the government have instilled “alarmist“ rhetoric in some of the lawmakers. He says that if the law focuses too much on preserving business opportunities for companies, the result may be similar to how the current tech industry enjoys almost unrestricted access to their customers’ personal information.
What's more, it will be a difficult challenge to enact the law across all 27 EU nations. The law has not made it clear which national institute will be in charge of oversight.
An official with a leading role in the implementation of the AI Act anticipates a fight between local telecoms and competition and data protection watchdogs over “who gets to call the shots“. “It could get messy,“ the official says. “There is a disparity of perspectives over who should be the enforcer."
In other words, when AI companies launch their products across the EU, they may run into different regulatory standards in different member countries.
The EU had hoped that the creation of its AI Office would help mitigate some of the problems. But recruitment has been sluggish, with 140 full-time positions yet to be filled. This is because policy experts and technical staff are hard to come by. Many of them may already have been hired by big tech companies.
It's also worth noting that the EU's regulatory quagmire may not be confined to Europe. The U.K., U.S., and other G7 nations are looking to introduce their own safeguards on AI. The EU's troubles may serve as a teachable moment. But it remains to be seen if other governments will have any more luck finding the perfect balance between effective protection and ample room for innovation.
Source： EU, MIT, FT, CNBC (2024-07-19)
About Us
InvesTaiwan InvesTaiwan Service Center Department of Investment Promotion, MOEA Department of Investment Review, MOEA Contact TaiwanWhy Taiwan
Reasons to Invest Overview Key Industries for Investment Promotion Incentives Important Policies Success Stories
MOEA
MOEA DOIP
MOEA YouTube
Let's Juiker Together
