CoChat
Sign Up
Cited by 255Ethics and Social Impacts of AIArtificial Intelligence in LawPolicing Practices and Perceptions

The DOI matches its Crossref record; 62 of the 62 references it cites were found in a registry or an index.

Artificial Intelligence, Predictive Policing, and Risk Assessment for Law Enforcement

Berk, Richard A.·2021·Annual Review of Criminology
DOI 10.1146/annurev-criminol-051520-012342
Open this paper in CoChat — free account

Abstract

There are widespread concerns about the use of artificial intelligence in law enforcement. Predictive policing and risk assessment are salient examples. Worries include the accuracy of forecasts that guide both activities, the prospect of bias, and an apparent lack of operational transparency. Nearly breathless media coverage of artificial intelligence helps shape the narrative. In this review, we address these issues by first unpacking depictions of artificial intelligence. Its use in predictive policing to forecast crimes in time and space is largely an exercise in spatial statistics that in principle can make policing more effective and more surgical. Its use in criminal justice risk assessment to forecast who will commit crimes is largely an exercise in adaptive, nonparametric regression. It can in principle allow law enforcement agencies to better provide for public safety with the least restrictive means necessary, which can mean far less use of incarceration. None of this is mysterious. Nevertheless, concerns about accuracy, fairness, and transparency are real, and there are tradeoffs between them for which there can be no technical fix. You can't have it all. Solutions will be found through political and legislative processes achieving an acceptable balance between competing priorities.

Overview

Integrity signals

Registry

Crossref · 10.1146/annurev-criminol-051520-012342

Cites

62 of 62 references verified0 citation errors · 0 not checkable

Statistics

Recomputing p-values needs the paper’s full text.

Verification

Claims the paper makes

This paper’s full text isn’t openly available, so its claims and self-check weren’t run. The registry and citation checks still apply.

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?

Citations

Cites

Show all 62 references

Research works printed but not linked: 44

  1. Ackerman E., 2017, IEEE Spectrum
  2. Aldor-Noiman S, 2016, Stat. Sin., 26, 1587
  3. Angwin A, 2016, ProPublica
  4. Anselin L, 2000, Criminal Justice 2000: Measurement and Analysis of Crime and Justice, 213
  5. Berk RA., 2018, Machine Learning Forecasts of Risk in Criminal Justice Settings
  6. Berk RA, 2013, J. Criminol. Public Policy, 12, 515
  7. Berk RA, 2016, J. Empir. Legal Stud., 13, 95
  8. Burgess EM., 1928, The Working of the Indeterminate Sentence Law and the Parole System in Illinois, 205
  9. Caplan JM., 2011, J. Policy Dev. Res., 13, 57
  10. Cellen-Jones R., 2014, BBC News: Technology
  11. Chinoy S., 2019, New York Times
  12. Coglianese C, 2017, Georget. Law J., 105, 1147
  13. Corbett-Davies S, 2018, The measure and mismeasure of fairness: a critical review of fair machine learning
  14. Goel S, 2019, Risk-Resil. Res.
  15. Goodfellow I, 2016, Deep Learning
  16. Groff ER, 2004, Crime Prev. Stud., 13, 29
  17. Hamilton M., 2016, Am. Crim. Law Rev., 52, 231
  18. Helmstetter A, 2003, J. Geophys. Res., 108, 2483
  19. Hill K., 2020, New York Times, 3
  20. Hollywood JS, 2019, Real-Time Crime Centers in Chicago
  21. Huq AZ., 2019, Duke Law J, 68, 1043
  22. Johnson SD, 2007, Prospective crime mapping in operational context
  23. Kearns M, Neel S, Roth A, Wu Z. 2018b. Preventing fairness gerrymandering: auditing and learning subgroup fairness. arXiv 1711.05144v4 [cs.LG] · arXiv
  24. Kessel JM., 2019, New York Times
  25. Kroll JA, 2017, Univ. Pa. Law Rev., 165, 633
  26. Lyngstad T, 2017, Predicting future crime at birth
  27. Malkevitch J., 2007, Taxi! American Mathematical Society
  28. Mayson SG., 2019, Yale Law J, 128, 2218
  29. Metz C., 2019, New York Times
  30. Mitchell M., 2019, New York Times
  31. Mullainathan S., 2019, New York Times
  32. Pande V., 2018, New York Times
  33. Pascanu R, Gulcehere C, Cho K, Begio Y. 2014. How to construct deep recurrent neural networks. arXiv:13126026v5 [cs.NE]
  34. Pindyck RS, Rubinfeld DL. 1981. Econometric Models and Economic Forecasts . New York: McGraw Hill. 2nd ed.
  35. Poole D, 1998, Computational Intelligence: A Logic Approach
  36. Rachlinski JJ, 2009, Notre Dame Law Rev, 84, 1195
  37. Reiss AJ., 1951, Am. J. Sociol., 17, 268
  38. Rosser G, 2016, Appl. Spat. Anal., 12, 5
  39. Smith CS., 2019, New York Times
  40. Spielkamp M., 2017, MIT Technology Review
  41. Starr S., 2014, New York Times
  42. Wang B, Yin P, Bertozzi AL, Brantingham PJ, Osher SJ, Xin J. 2017a. Deep learning for real-time crime forecasting and its ternarization. arXiv:1711.08833v1 [cs.LG] · arXiv
  43. Wang B, Zhang D, Zhang D, Brantingham PJ, Bettozzi AL. 2017b. Deep learning for real time crime forecasting. arXiv:1707.03340v1 [math.NA] · arXiv
  44. Weisburd D, 2004, The growth of Compstat in American policing

Cited by

Sorted by citations

Related

See all

Metadata

Authors
Berk, R. A.
Affiliations
University of Pennsylvania (US)
Corresponding author
Berk, Richard A.
ORCID
Berk, Richard A.
Venue
Annual Review of Criminology
Publisher
Annual Reviews
Published
2021-01-13
Volume
4
Issue
1
Pages
209-237
ISSN
2572-4568 (ISSN-L)
DOI
10.1146/annurev-criminol-051520-012342
OpenAlex
W3100801413
Cited by
213
References
62
Citations by year
2026: 73 · 2025: 50 · 2024: 37 · 2023: 23 · 2022: 25 · 2021: 4
FWCI
2.55
Type
journal article
Language
en
Open access
closed
Locations
Annual Review of Criminology (publishedVersion)
Topics
  • Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
  • Social Sciences › Law › Artificial Intelligence in Law
  • Social Sciences › Political Science and International Relations › Policing Practices and Perceptions
Keywords
artificial intelligence, predictive policing, algorithmic bias, algorithmic transparency, algorithmic fairness, crime prediction, nonparametric regression, public safety, law enforcement

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