Can the European Union stop artificial intelligence?

   A European Union plan to regulate artificial intelligence could lead companies to violate proposed rules of mass surveillance and discrimination fined millions of euros. The bill, leaked before its official release later this month, shows the EU is trying to figure out "a third way" to regulate AI, between the free market of the US and authoritarian China.

   As stated before, the rules will ban AI designed to manipulate people to "harm them", conduct indiscriminate surveillance, or calculate "social scores". Much of the language is ambiguous enough that regulations can cover the entire advertising industry or nothing at all. In all cases, the military and any agency that guarantees public security are exempt.
   A number of “high-risk” activities will be allowed, subject to tight control, including measures to prevent the inclusion of racial, gender or age biases in AI systems. To the possible goal, the law refers to systems for automating vacancies, assigning positions at school, college or university, measuring credit scores or deciding application results. apply for a visa. Violating companies can be fined up to 20 million euros, or 4% of global sales.
   In a way, the news comes as no surprise that the president of the European Commission, Ursula von der Leyen, has promised to urgently introduce AI legislation when she is elected in 2019. But Lilian Edwards at Newcastle University , UK, said the bill. will be related to the technology industry. "I welcome that expectation from you, but you can't imagine how it will play out in this state," she said.
Edwards began comparing the approach to the way that the EU requires consumer products needed to meet certain requirements to be imported. “It's a lot harder to do with AI because it's not always a simple product,” she said. "You are headed for a trade war with global Silicon Valley or poor enforcement."
   China and the US have made great strides in the deployment of AI in a wide range of industries, including national security and law enforcement. In China, citizens' daily movements in many cities are tracked with facial recognition, and more public and private trials of "social credit scores" will eventually be rolled out across the country. country. . This score can be lowered for a violation like playing a computer game for too long or crossing the street at a red light for pedestrians and can be raised by donating to charity. . If your score is too low, you could be denied a rail ride or embarrassed on the online listing.
   Meanwhile, in the US, where many of the tech giants are headquartered, the Donald Trump administration encourages a free, free market approach, while the incumbent Joe Biden does not take a public stance. steady.
   Daniel Leufer at Access Now, one of the groups that previously advised the EU on AI, said Europe has long had a strategy for taking a third approach between the US and China in terms of technology regulation and for That the draft has great promise.
But he warns there are "big signs" around some of the bill's elements, such as the establishment of the European Ministry of Artificial Intelligence. "They will have a huge influence on what is added or taken off the high risk and banned list," which means exactly who will be key.
   The EU has had earlier success in influencing global technology policy. Its general data protection regulations, introduced in 2018, have inspired similar laws in non-EU countries and in California, the home town of Silicon Valley. In response, however, some US companies simply block EU customers from accessing their services.
   It remains to be seen whether the UK will follow the EU in regulating AI as it leaves the bloc. The UK's Department for Business, Energy & Industrial Strategy told New Scientist that the government has set up an independent panel called the Regulatory Horizon Council to advise on what regulations are needed to react with new technologies like AI.