The European Commission (EC) proposed new rules covering the impact of AI on humans and businesses, targeting creation of a trustworthy environment for the development of innovative products and services in the European Union (EU).
In its statement, the EC explained rules would be founded on a risk-based approach to AI, and seek to find a balance between promoting uptake of the technology with protecting people.
It plans to ban AI systems considered to pose “a clear threat” to safety, livelihoods and rights of people, including services which manipulate human behaviour “to circumvent users’ free will” or allow government-developed social credit systems.
AI systems deemed high risk include those covering employment; education; law enforcement; and critical infrastructure, for example transport. These will be subjected to “strict obligations” before appearing on the market.
The EC emphasised systems for biometric identification will also face strict requirements, with their use in public spaces banned save for “narrow exceptions” such as investigating terrorist threats or cases of missing children.
It noted the “vast majority” of AI is expected to be low-risk citing email spam filters and video games as example.
EC EVP Margrethe Vestager said the regulations will be “future-proof and innovation-friendly”.
Thierry Breton Commissioner for the Internal Market, added the proposed rules “aim to strengthen Europe’s position as a global hub of excellence in AI from the lab to the market”, and help to “harness the potential of AI for industrial use”.
The proposed rules must be ratified by the European Parliament and EU member states before they can be adopted.
In 2020, the EC started a public consultation on the proposals, claiming at the time it intended to create incentives for accelerating AI adoption across the EU.