[Turing-Southampton] Request for feedback on Turing-ICO draft guidance on ‘Explaining decisions made with AI’

[Susan Davies]

The Turing has been working together with the Information Commissioner’s Office over the last year to produce draft guidance on ‘explaining decisions made with AI’. This came about through the recognition that organisations are increasingly using AI to help them make decisions about people, and that the implications for individuals and society are potentially significant.

We were tasked to provide guidance on how to explain these decisions by the government’s Office for AI in the AI Sector Deal. This guidance is now out for consultation until 24th January 2020.

The guidance is aimed at organisations that are using AI on personal data. It is split into three parts:

Part 1: The basics of explaining AI defines the key concepts and outlines a number of different types of explanation. Four core principles inform the overarching approach that you should take to ensure the fully automated or AI-enabled decisions you make can be appropriately explained to individuals. This part also sets out the applicable legal framework, and discusses some relevant benefits and risks. Everyone will find this useful, but particularly your Data Protection Officer and compliance team.

Part 2: Explaining AI in practice helps you with the practicalities of explaining these decisions and providing explanations to individuals. It shows how you go about selecting the appropriate explanation for your sector and use case, how to choose an appropriately explainable model, and which tools you can use to extract an explanation from less interpretable models. It also helps you to turn the output you get from your AI model into an understandable reason for the decision recipient. This part will primarily be helpful for technical teams in your organisation, although your Data Protection Officer and compliance team will also find it helpful.

Part 3: What explaining AI means for your organisation goes into the various roles, policies, procedures and documentation that you can put in place to ensure your organisation is set up to provide meaningful explanations to affected individuals. This part is primarily aimed at your organisation’s senior management team, however your Data Protection Officer and compliance team will also find it useful.

We would greatly appreciate your feedback on the technical aspects of this guidance (particularly Part 2 on how to explain AI models), but of course welcome comments on any part of the guidance. The link for providing feedback and finding more information on the consultation is: https://ico.org.uk/about-the-ico/ico-and-stakeholder-consultations/ico-and-the-turing-consultation-on-explaining-ai-decisions-guidance/<https://eur03.safelinks.protection.outlook.com/?url=https%3A%2F%2Fico.org.uk%2Fabout-the-ico%2Fico-and-stakeholder-consultations%2Fico-and-the-turing-consultation-on-explaining-ai-decisions-guidance%2F&data=01%7C01%7C%7C00c6bc3c45aa4950e0ca08d782211556%7C4a5378f929f44d3ebe89669d03ada9d8%7C0&sdata=De0Y6Vuh2elh7BJrHz3oPVu%2BbOdSMMC0Q3Cf%2FKh1wOc%3D&reserved=0>

Please contact explain at ico.org.uk<mailto:explain at ico.org.uk> if you have any questions.


_____________________________________________
Susan Davies
Coordination Manager, Web Science Institute<https://www.southampton.ac.uk/wsi/index.page?>
University Liaison Manager, The Alan Turing Institute<https://www.southampton.ac.uk/wsi/alan-turing-institute/alan-turing-institute.page>
Room 3041, Building 32
Web Science Institute
University of Southampton
Southampton SO17 1BJ

023 8059 3523 | 07768 266464


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ecs.soton.ac.uk/pipermail/turing-southampton/attachments/20191216/235e3d73/attachment.html