Sustainable AI

Why we need sustainable AI

Several concepts have been advanced to help the public sector manage AI risks, such as explainability, ethics, responsibilty, trust, international human rights law or algorithmic self-regulation. All of these fall short when applied. They require a critical and engaged public and put significant demands on the capacity of the public sector. Regulation and implementation of AI cannot be meaningfully guided if regulators and impementors do themselves not understand the platforms at issue or the ways they interact with human agents.

Public sector actors managing the regulation and implementation of AI have unique obligations and capacity constraints compared to the private sector and currently lack any operational framework with which to manage them.

The concept of Sustainable AI meets precisely these needs. Firstly, the notion of sustainability has been clearly defined in the context of sustainable development. This has the added value of policy recognition and alignment for the public sector, which is generally already committed to meeting the Sustainable development Goals. Secondly, it is a systemic concept, covering all aspects of social good and responsibility and, most importantly, it captures the particular relationship between public sector agencies and the publics they serve.

Besides research and writing together, we organise a Sustainable AI Lab in which we design and develop participatory processes that strengthen the adoption and implementation of AI in the public sector as well as people's interaction with public services employing AI.

Funding

We are actively looking for funding opportunities to give our collaboration a more permanent character. Please contact Christopher Wilson, our project coordinator, for more information.