About

In 2021, the European Commission (EC) proposed regulation for AI systems, outlining a risk-based approach to governing innovation and application of AI systems. The proposed regulation categorizes AI systems based on the associated risks to health, safety, and fundamental rights of people. The proposed regulation suggests AI systems be allocated a risk category of “unacceptable risk”, “high risk”, “minimal risk” or “no risk”. In addition, the draft regulation outlines a list of domains of AI use cases that would be considered high-risk and thus subject to requirements outlined in the regulation.

While the proposed regulation is widely considered as a key step in the right direction, there are several open questions when it comes to applying the regulation and understanding what could be considered high-risk under the regulation. This has also been a particular focus of a lot of industry feedback received by the European Commission.

One of the best mechanisms to support the rollout of the EC proposed regulation and other AI regulations is to offer concrete use cases in an easily accessible central repository of real-world use cases to help inform and improve policy and technological outcomes. This work would also close a gap between policy considerations and risk-based expectations and outcomes and provide a level of clarity that can enable trustworthy AI innovations and implementations.

This Industry Connections initiative will identify, aggregate, and classify AI use cases across key sectors, with an initial focus on highest categories of risk as defined by the proposed EC regulation. We aim to conduct an open call for use cases submissions; vet and assess associated risk tiers using the EC’s definition, expertise of IC18-004 ECPAIS members, as well as of this initiative’s expert volunteers; curate use cases with proper abstract and detailed implications; and where possible, invite submitters to share approaches to detailing the potential positive and negative impacts, and to mitigating the risks and potential harms. We intend to conduct an open content review period to elicit feedback from external entities and experts. As this information is contextualized and standardized in form, the output of this work will be a searchable database that can be used by developers, deployers and policymakers. In addition, a report will be created to release key learnings and describe how the work may be further built upon. At the conclusion of this initiative, we anticipate hosting a webinar on the identified use cases.