|
EU general approach for 'trustworthy' artificial intelligence
The European Union’s top AI experts are set to present a list of 33 recommendations on how to boost a European AI industry to the bloc’s leaders next Wednesday.
Europe needs carefully targeted policy-making to catch up in the global race for artificial intelligence, according to AI experts. The 52 top AI specialist advise against a one-size-fits-all approach for AI regulation, but call for a review of the bloc’s existing rules, as well as for defining “red lines” and discussing the potential ban of high-risk AI applications “with affected stakeholders.”
The document urges the lawmakers to make sure AI is not used for the mass surveillance of individuals, either by governments or private actors. “AI enabled mass scale scoring of individuals” should be banned, the expert group demands, and there need to be “very clear and strict rules for surveillance for national security purposes and other purposes claimed to be in the public or national interest.”
The experts also recommend policy measures to boost Europe’s AI industry, such as public funding for businesses that use “trustworthy AI.”
Guiding principles vs hard rules
The experts’ recommendations aim defining the conditions under which AI should be developed and deployed in the EU internal market. According to the group, AI does need monitoring, but regulation should come as broad guiding principles rather than specific detailed rules stating out what AI can and can’t do. The document wants to make sure that legislation does not burden innovation.
As a first step, they recommend to map the already existing legislations and assess them according to AI systems. When it comes to data protection, for example, the EU should “consider whether existing laws allow sufficient access to public data and data for legitimate research purposes whilst preserving privacy and personal data protection.”
The experts recommend “a mandatory obligation to conduct a trustworthy AI assessment” for certain high-risk AI applications developed by the private sector. They suggest creating de-regulated sandbox environments, establishing auditing mechanisms and involving civil society to define “red lines” for which applications should be banned.
No robot personhood
Europe’s 52 top AI specialists take a clear stand against a controversial suggestion that robots should legally be considered “persons” to hold them accountable for their actions. This approach is fundamentally inconsistent with the principle of human agency, accountability and responsibility, and to pose a significant moral hazard.
However, an update of the EU’s legal framework is needed to enable the development of cutting-edge AI applications, AI will indeed boost the EU economy and the continent lags behind the global champion – The US and China.
A public-private partnership to make EU “Trustworthy” AI an asset to compete on global market
The document recommends the setting-up of a public-private partnership for the EU AI industry that will likely influence policy decisions made by the next European Commission.
These recommendations take roots in the previously issued document entitled “trustworthy artificial intelligence”. EU wants the development of AI which adheres to high ethical standards. According to the experts. This specificity will turn into a competitive advantage for European manufacturers, with consumers eventually preferring them over those produced by competitors in China or the U.S.
Instead of trying to catch up in the field of consumer applications, where U.S. and Chinese companies dominate, Europe should focus on developing business-to-business solutions as well as “public-to-citizens” services. |
