An Accountability for Algorithms Act? Proposals from the Institute for the Future of Work

The Institute for the Future of Work (IFOW) has proposed an Accountability for Algorithms Act because, amongst other things, existing statutory and regulatory frameworks have not kept pace with developments in what the technology can do or how it is used.

This comes at the same time as

EU adopts proposals on how the EU can regulate and legislate for Artificial Intelligence, and is likely to form part of a wider debate on how artificial intelligence should be regulated and legislated for.

The IFOW's proposals are for an act which:

- introduces new corporate duties, emphasising pre-emptive governance and action to evaluate AI's equality impacts and to make adjustments to eliminate unlawful discrimination;

- introduces a mandatory transparency obligation to record facts, purposes and outcomes of algorithmically-assisted decision-making, a right to know about the same, and a right for workers to be involved to a reasonable extent on the development of such systems;

- a new regulatory forum to help enforce these obligations and co-ordinate regulatory responses (e.g. from the Information Commissioner's Office and Human Rights Commission).

there is a regulatory lacuna: existing statutory and regulatory frameworks do not provide sufficient accountability for the key individuals at the critical stage in the design and use of data-driven technologies

https://uploads-ssl.webflow.com/5f57d40eb1c2ef22d8a8ca7e/5f994c35cac984f4cd9bfb43_IFOW%20-%20An%20Accountability%20for%20Algorithms%20Act%20-%20Executive%20Briefing%20(28.10.20)%20(1).pdf