CoChat

AI for the Common Good?! Pitfalls, challenges, and ethics pen-testing

Open access
Berendt, Bettina · 2019 · Paladyn, Journal of Behavioral Robotics

The DOI matches its Crossref record; 51 of the 52 references it cites were found in a registry or an index; 10 of the 56 sources its claims cite were checked for support.

Abstract

Recently, many AI researchers and practitioners have embarked on research visions that involve doing AI for “Good”. One frequent theme in current ethical guidelines is the requirement that AI be good for all, or contribute to the common good. The paper proposes ethics pen-testing as a method for making impediments visible and attenuating them.

Overview

CoChat Summary

## TL;DR

The paper argues that “AI for the Common Good” is impeded by four recurring issues: how problems are framed, who defines them, how knowledge is treated, and how side effects and dynamics are handled. It proposes “ethics pen-testing” as an adversarial design method to surface these weaknesses and improve AI systems.

## Key findings

  • The author’s exploratory survey covered 99 contributions across four venues: 24 from Data Science for Social Good 2017, 4 from SoGood 2017, 15 from the 2017 AAAI Spring Symposium on AI for Social Good, and 56 from the 2nd AI for Good Global Summit.
  • In the surveyed material, the vast majority of contributions used one social goal and one computational goal; there was generally no attempt at alternative framings of the social problem, and the difference between problem versions (Q1') was not the topic of any paper.
  • For stakeholder involvement, the large majority of contributions took the problem definition from a non-academic project partner; only one contribution described an explicit multistakeholder process, and another mentioned such a process followed by partners.
  • On dynamics, only one contribution in the sample included an explicit consideration of possible future dynamics; the author says the other Q1–Q4 considerations are not an integral part of current operational research practices in AI/DS for Social Good.
  • The paper reports that in the first three venues, 8 out of 43 contributions covered SDG #3 (Good Health and Well-being); 5 could be linked to SDG #11 (Sustainable Cities and Communities), and 2 to transportation. In the AI for Good Global Summit, 11 contributions covered SDG #3, and one track had 13 contributions on Smart Cities and another had 6 on AI and Satellite Imagery.

## Methods

  • Conceptual/normative argument: the paper draws on AI ethics codes, political philosophy, and computer-science ideas to derive four lead questions about Common Good-oriented AI.
  • Exploratory document survey: it reads and assesses all hyperlinked contributions from four recent AI/DS-for-good venues, coding whether they explicitly addressed the four questions.
  • The survey examined content and whether contributions contained explicit information relating to Q1–Q4; the author notes the coding was exploratory and simple.

## Limitations

  • The paper says the conference survey was exploratory, the coding exercise was simple, and future work will use a codebook and more coders.
  • The author notes that the findings may be an artifact of publication conventions that favor unambiguity and a linear research story, and that many surveyed documents were very short; this may have hidden alternatives and process details.
  • As an evident limitation from the text: the survey covers only selected conference materials from four venues, so it is not a comprehensive study of AI for the Common Good.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1515/pjbr-2019-0004

Cites

51 of 51 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Verification

Claims the paper makes

8 Valid · 9 Not valid · 31 Not checked · 90 Own results and interpretations

28 more claims

Create a free account to see all 138 claims and to verify the 16 references still open.

See all claims — free account
  • It may state two separate claims as one. 41
  • Probably not a substantial claim. 93
  • The quote only partly states this claim. 1

Create a free account to see these claims.

Citations

Cites

Show all 51 references

Research works printed but not linked: 18

  1. [3] ACM Code of Ethics and Professional Conduct, 1992, https://www.acm.org/about-acm/acm-code-of-ethics-andprofessional-conduct
  2. [5] The IEEE Global Initiative for Ethical Considerations in Artificial Intelligence and Autonomous Systems, Ethically Aligned Design: A Vision for Prioritizing Human Wellbeing with Artificial Intelligence and Autonomous Systems, Version 1 for public discussion, 2016, http://standards.ieee.org/develop/indconn/ec/ead_v1.pdf
  3. [16] B. Mols, AI for the Common Good, ACM News, Jun. 20, 2017, https://cacm.acm.org/news/218696-ai-for-the-commongood/fulltext
  4. [17] A. Tanweer, B. Fiore-Gartland, Cross-sector collaboration in Data Science for Social Good: Opportunities, challenges, and open questions raised by working with academic researchers, In: Data Science for Social Good Conference, Sep. 28-29, 2017, Chicago, IL, http://dssg.uchicago.edu/wp-content/uploads/2017/09/tanweer.pdf
  5. [22] S. J. Russell, P. Norvig, Artificial Intelligence: A Modern Approach, Prentice Hall, Englewood Cliffs, NJ, 1995
  6. [49] T. T. Nguyen, P.-M. Hui, F. M. Harper, L. Terveen, J. A. Konstan, J. A., Exploring the filter bubble: The effect of using recommender systems on content diversity, In: Proceedings of International World Wide Web Conference Committee (WWW’14), ACM, New York, 201410.1145/2566486.2568012 · DOI · Paper page
  7. [57] G. Rockwell, B. Berendt, Information wants to be free, or does it? The ethics of datafication, Electronic Book Review, 2017, http://electronicbookreview.com/thread/technocapitalism/datafiction
  8. [67] R. Gavaldà, I. Koprinska, S. Kramer (Eds.), Proceedings of the Second Workshop on Data Science for Social Good co-located with European Conference on Machine Learning and Principles and Practice of Knowledge Dicovery in Databases (ECML-PKDD 2017), Skopje, Macedonia, Sep. 18, 2017, CEUR Workshop Proceedings Vol-1960, http://ceur-ws.org/Vol-1960/
  9. [68] E. Horvitz, B. Grosz, A. Greenwald, D. Parkes, C. Gomes, S. Smith, et al. (Eds.), Artificial Intelligence for the Social Good, Papers from the 2017 AAAI Spring Symposium, Technical Report SS-17-01, AAAI, Palo Alto, CA, 2017, https://www.aaai.org/Library/Symposia/Spring/ss17-01.php
  10. [69] J. Garriga, J. Piera, F. Bartumeus, A Bayesian framework for reputation in citizen science, In: Proceedings of the Second Workshop on Data Science for Social Good, CEUR Workshop Proceedings, 2017, 1960, 1-18, http://ceur-ws.org/Vol-1960/paper6.pdf
  11. [71] J. Auerbach, H. Barton, T. Blunt, V. Chaganti, B. Ghai, A. Meng, et al., Using data science as a community advocacy tool to promote equity in urban renewal programs: An analysis of Atlanta’s Anti-Displacement Tax Fund, In: Data Science for Social Good Conference 2017, Sep. 28-29, 2017, Chicago, IL, http://dssg.uchicago.edu/wp-content/uploads/2017/09/auerbach.pdf
  12. [73] M. Prasad, Back to the future: A framework for modelling altruistic intelligence explosions, In: AAAI 2017 Spring Symposium on Artificial Intelligence for the Social Good, https://aaai.org/ocs/index.php/SSS/SSS17/paper/view/15326
  13. [74] Y.-H. Wang, Y.-Y. Chen, S.-C. Chen, C.-K. Liu, T. C. Hsieh, Data for Social Good: A case study of building an effective public-private partnership on domestic violence prevention, In: Data Science for Social Good Conference 2017, Sep. 28- 29, 2017, Chicago, IL, http://dssg.uchicago.edu/wp-content/uploads/2017/09/hsieh.pdf
  14. [75] A. Delaunay, J. Guérin, Wandering detection within an embedded system for Alzheimer suffering patients, In: AAAI 2017 Spring Symposium on Artificial Intelligence for the Social Good, https://aaai.org/ocs/index.php/SSS/SSS17/paper/view/15317
  15. [76] O. Bendel, LADYBIRD: The animal-friendly robot vacuum cleaner, In: AAAI 2017 Spring Symposium on Artificial Intelligence for the Social Good, https://aaai.org/ocs/index.php/SSS/SSS17/paper/view/15277
  16. [77] P. Baumgartner, Challenges in assessing predictive bias, In: Data Science for Social Good Conference 2017, Sep. 28- 29, 2017, Chicago, IL, http://dssg.uchicago.edu/wp-content/uploads/2017/09/baumgartner.pdf
  17. [82] C.-K. Liu, T. C. Hsieh, Lessons learned from using data science to empower change agents across data silos, In: Data Science for Social Good Conference 2017, Sep. 28-29, 2017, Chicago, IL, https://dssg.uchicago.edu/wp-content/uploads/2017/09/liu.pdf
  18. [84] H. Holmestad, Predicting risk of long-term unemployment, In: Data Science for Social Good Conference 2017, Sep. 28-29, 2017, Chicago, IL, https://dssg.uchicago.edu/wp-content/uploads/2017/09/holmestad.pdf

Cited by

See all 23

Metadata

Authors
Berendt, B.
Affiliations
KU Leuven (BE)
Corresponding author
Berendt, Bettina
Venue
Paladyn, Journal of Behavioral Robotics
Published
2019-01-01
Volume
10
Issue
1
Pages
44-65
ISSN
  • 2080-9778 (ISSN-L)
  • 2081-4836
OpenAlex
W2898730196
Cited by
8
References
52
Citations by year
2026: 1 · 2025: 1 · 2023: 3 · 2021: 1 · 2020: 2
FWCI
1.31
Type
journal article
Language
en
Open access
Locations
  • Paladyn Journal of Behavioral Robotics (publishedVersion, cc-by)
  • Lirias (acceptedVersion, other-oa)
  • arXiv (Cornell University) (submittedVersion)
  • arXiv (Cornell University)
Topics
  • Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
  • Social Sciences › Sociology and Political Science › Innovation, Technology, and Society
  • Physical Sciences › Computer Science › Artificial Intelligence › Artificial Intelligence Applications
Keywords
AI ethics, AI for social good, problem formulation, role of knowledge, AI design, moral norms

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.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

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

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: AI for the Common Good?! Pitfalls, challenges, and ethics pen-testing. https://app.cochat.ai/papers/10.1515/pjbr-2019-0004
The paper
Berendt, B. (2019). AI for the Common Good?! Pitfalls, challenges, and ethics pen-testing. Paladyn, Journal of Behavioral Robotics. https://doi.org/10.1515/pjbr-2019-0004

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.