The Commission outlines various approaches needed for dealing with AI systems based on the risks posed them being unacceptable, high, limited and minimal
The European Commission (EC) has proposed new rules and actions pertaining to the use of artificial intelligence (AI), under which certain “unacceptable” uses of the technology will be banned.
The new rules and actions are part of a regulatory framework on the technology along with a revised coordinated plan. These are expected to promote the development of AI and at the same time address the possible high risks it has to safety and fundamental rights.
EC Europe fit for the Digital Age Executive Vice-President Margrethe Vestager said: “On Artificial Intelligence, trust is a must, not a nice to have. With these landmark rules, the EU is spearheading the development of new global norms to make sure AI can be trusted.
“By setting the standards, we can pave the way to ethical technology worldwide and ensure that the EU remains competitive along the way. Future-proof and innovation-friendly, our rules will intervene where strictly needed: when the safety and fundamental rights of EU citizens are at stake.”
According to the Commission, AI systems that are thought to pose a clear threat to people’s safety, their livelihoods and rights will be banned. Included in these are AI-powered systems or applications that control human behaviour to bypass users’ free will.
Also banned will be systems that enable governments to undertake social scoring.
On the other hand, all high-risk AI systems will be put to strict obligations prior to their launch in the market, said the Commission.
Some of the AI systems identified as high-risk include the use of the technology in critical infrastructures such as transport, educational or vocational training, law enforcement, and essential private and public services, among others.
The EC emphasised that all remote biometric identification systems are put under the high risk category and their use will be subject to strict requirements. The Commission said that their live uses are banned in principle in publicly accessible spaces for the purpose of law enforcement.
For AI systems like chatbots and others that pose limited risk, specific transparency obligations have to be fulfilled.
The EC said that its legal proposal doesn’t have any restrictions on the use of AI-enabled video games, spam filters, or other such applications, which come under the minimal risk category.
The next steps would need the European Parliament and the Member States to adopt the proposals on a European approach for AI and also on machinery products. After they are adopted, the regulations will be applicable directly across all the member states of the European Union.