CoChat
Sign Up
Cited by 126Ethics and Social Impacts of AIE-Government and Public ServicesDigitalization, Law, and Regulation

Understanding artificial intelligence ethics and safety: A guide for the responsible design and implementation of AI systems in the public sector

Open access
Leslie, David·2019·Zenodo
DOI 10.5281/zenodo.3240529
Open this paper in CoChat — free account

Abstract

A remarkable time of human promise has been ushered in by the convergence of the ever-expanding availability of big data, the soaring speed and stretch of cloud computing platforms, and the advancement of increasingly sophisticated machine learning algorithms. Innovations in AI are already leaving a mark on government, by improving the provision of essential social goods and services from healthcare, education, and transportation to food supply, energy, and environmental management. These bounties are likely just the start. The prospect that progress in AI will help government to confront some of its most urgent challenges is exciting, but legitimate worries abound. As with any new and rapidly evolving technology, a steep learning curve means that mistakes and miscalculations will be made and that both unanticipated and harmful impacts will occur. In order to manage these impacts responsibly and to direct the development of AI systems toward optimal public benefit, The Alan Turing Institute's public policy programme partnered with the Office for Artificial Intelligence and the Government Digital Service to produce guidance on the responsible design and implementation of AI systems in the public sector. The guide, Understanding Artificial Intelligence Ethics and Safety, is the most comprehensive guidance on the topic of AI ethics and safety in the public sector to date. It identifies the potential harms caused by AI systems and proposes concrete, operationalisable measures to counteract them. The guide stresses that public sector organisations can anticipate and prevent these potential harms by stewarding a culture of responsible innovation and by putting in place governance processes that support the design and implementation of ethical, fair, and safe AI systems. The guidance is relevant to everyone involved in the design, production, and deployment of a public sector AI project: from data scientists and data engineers to domain experts, delivery managers and departmental leads. Our aim -- and hope -- in writing the guide is to encourage civil servants interested in conducting AI projects to make considerations of AI ethics and safety a first priority.

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

DataCite · 10.5281/zenodo.3240529

Cites

0 citation errors · 0 not checkable

Statistics

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • The study in [383] suggests that the incorporation and consideration of explainability in practical AI design and deployment workflows should comprise four major methodological steps.

    “Among them, we pause at the thorough study in [383], which suggests that the incorporation and consideration of explainability in practical AI design and deployment workflows should comprise four major methodological steps:”

    From Explainable Artificial Intelligence (XAI): Concepts, taxonomies, opportunities and challenges toward responsible AI (2020) · cites this paper as [383]

    The guide’s introductory material (page not available) says it provides end-to-end guidance on ethical AI design and implementation. The available passages do not address explainability or identify four methodological steps; searches for those terms returned only introductory material. The specific attribution therefore cannot be verified from our copy. (LLM, conf=0.72)

    Experimental — accuracy not yet measured

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Cites

The publisher hasn’t deposited this paper’s reference list.

Cited by

Sorted by citations

Related

Authors
Leslie, D.
Corresponding author
Leslie, David
ORCID
Leslie, David
Venue
Zenodo
Publisher
Zenodo
Published
2019-06-11
DOI
10.5281/zenodo.3240529
OpenAlex
W6931119994
Cited by
126
References
0
Citations by year
2026: 31 · 2025: 21 · 2024: 7 · 2023: 11 · 2022: 5 · 2021: 38 · 2020: 10 · 2019: 3
Type
report
Language
en
Open access
Licence
creativecommons.org/licenses/by-nc-sa/4.0/legalcode
Locations
  • University of North Texas Digital Library (University of North Texas) (submittedVersion)
  • Zenodo (CERN European Organization for Nuclear Research)
Funders
Government of the United Kingdom
Topics
  • Social Sciences › Safety Research › Ethics and Social Impacts of AI
  • Social Sciences › Political Science and International Relations › E-Government and Public Services
  • Social Sciences › Law › Digitalization, Law, and Regulation
Keywords
AI ethics, AI safety, public sector AI, responsible innovation, AI governance, ethical AI design, algorithmic fairness, digital public services, public sector employees

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review