Artificial Intelligence Applications

for Criminal Courts

An overview of artificial intelligence applications and associated
considerations for the criminal court system

This technology brief is the third in a four-part series that examines artificial intelligence (AI) applications in the criminal justice
system. This brief focuses on AI applications within criminal courts, with particular emphasis on AI’s role in addressing prosecutorial
needs. These AI applications and associated needs may also be relevant to other types of courts, such as traffic and civil courts, as
well as to other officers of the court—including defense counsel, judges, and court administrators.

It also introduces frameworks for evaluating AI applications and highlights critical risks to consider when deploying AI systems.
Although many of the examples in this brief have not yet been widely adopted, AI has the potential to address various needs within
the court system. Additional briefs include a high-level overview of AI within the criminal justice system and AI topics specifically
related to law enforcement and corrections.

Key Takeaways The courts play a critical role in the criminal justice system in ensuring the fair and
impartial administration of justice for all. As AI becomes more prevalent across society,
many criminal justice leaders are asking if AI-enabled technologies can help improve
¡ AI has the potential to transform
the court system. In other industries, AI has dramatically increased efficiency, expanded
many aspects of the court system
capabilities, and automated repetitive or mundane tasks. In the years ahead, AI will likely
in the years to come. Although
impact many aspects of the court system, including the prosecution and defense of
not yet ubiquitous, AI-enabled
crimes and the practice of law in both private and public service settings. This brief (1)
tools are already being used in
offers mental models for leaders in the criminal court system to use when evaluating AI
various applications relevant to
applications, (2) presents example AI applications and use cases, and (3) highlights key
the court system.
risk considerations within the criminal courts context.
¡ AI-enabled tools may address
This document explores AI within the criminal court system. Additional briefs address
pressing needs within the court specific application areas.
system—including managing
staffing and resources, processing
digital information, improving
AI in the Criminal
court operations, managing Justice System
cases, maintaining accountability,
and creating partnerships and
collaboration.
AI in
¡ AI systems that provide Law Enforcement
recommendations or predictions
in the context of the court system
should be approached with
caution and evaluated carefully. AI in Corrections

¡ Deploying AI-enabled tools

effectively requires investing in
a strategy for the operational,
procedural, and change Figure 1: Implementing AI impacts all stakeholders in the criminal justice community.
management efforts required for Briefs in this series frame AI within the community and focus on AI applications in law
successful implementation. enforcement, criminal courts, and corrections.

1
As the various actors in the criminal court system work towards justice, they face a growing set of challenges specific to
their roles. For example, prosecutors and defense attorneys must contend with the rapidly growing bodies of evidence
generated through modern technology while complying with a growing set of standards, such as those outlined in Brady
and Giglio.1 In addition, prosecutors and state-appointed public defenders in many states may face higher caseloads and
lower salaries than attorneys in other settings. These challenges and others are causing many states to have difficulty
finding and retaining new talent.2 This brief highlights examples of how AI can address these and other needs.

Court Needs that AI Might Address

Whereas the first brief in this four-part series explores the current state of AI within the criminal justice system broadly,
this brief discusses AI’s potential to address specific challenges that criminal courts face. Many of these challenges were
identified from NIJ’s Priority Criminal Justice Needs Initiative and the NIJ’s Courts Strategic Research Plan 2020–2024 as
illustrated in Figure 2.3, 4

Figure 2: Criminal court needs and challenges were identified through NIJ sponsored workshops and working groups with
practitioners.

1. U.S. Department of Justice. (2020). 9-5000: Issues related to trials and other court proceedings. Retrieved from https://www.justice.gov/jm/jm-9-5000-issues-related-trials-and-other-court-proceedings
2. Lawrence, D. S., Gourdet, C., Banks, D., Planty, M. G., Woods, D., & Jackson, B. A. (2019). Prosecutor priorities, challenges, and solutions. Santa Monica, CA: RAND Corporation. Retrieved from https://www.
rand.org/pubs/research_reports/RR2892.html
3. Lawrence, D. S., Gourdet, C., Banks, D., Planty, M. G., Woods, D., & Jackson, B. A. (2019). Prosecutor priorities, challenges, and solutions. Santa Monica, CA: RAND Corporation. Retrieved from https://www.
rand.org/pubs/research_reports/RR2892.html
4. National Institute of Justice. (2020). Courts Strategic Research Plan 2020–2024. (NCJ Number: 254684). Washington, DC: U.S. Department of Justice. Retrieved from https://nij.ojp.gov/library/
publications/courts-strategic-research-plan-2020-2024

Artificial Intelligence Applications

2 for Criminal Courts
An Introduction to AI

Defining Key Terms

AI can be broadly understood as a computer science domain that encompasses a wide range of technologies that
seek to mimic human intelligence. There are several different approaches to creating AI, and each has its strengths
and limitations. Recently, the explosion of available data combined with advances in machine learning and computing
hardware have paved the way for new AI applications and capabilities.

Improved AI algorithms have led to dramatic improvements in machine vision, natural language processing, robotic
process automation, and predictive analytics. Broadly speaking, today’s AI systems are much better at recognizing
patterns in data—including video/image data, text data, and numerical data—and using those patterns to classify objects
or make predictions.

Considerations for AI Success

AI has advanced in recent years; however, AI still has many technical and operational limitations. Data quality and
availability pose two of the biggest challenges to developing new AI applications. AI-generated predictions and
classifications depend on the data that are used to inform them, so AI outcomes will also reflect any data-related
limitations (e.g., missing data, unusable data, improperly stored data, or biased data). The process of developing (also
known as training) new AI systems often requires thousands of high-quality labeled examples for the algorithm to learn
from. Gathering, cleaning, and labeling data can be a time-intensive and expensive process. In addition, implementing
AI systems often requires process or behavioral changes that can be difficult or that can cause resistance from ecosystem
participants.

Even in the absence of technical or operational limitations, many AI applications raise ethical issues such as fairness,
transparency, accountability, privacy, and security. These concerns are particularly important for prosecutors and other
officers of the court because the fair and equitable pursuit of justice is a top priority.5 The first brief in this series includes
a list of key ethical questions that can serve as a starting point for criminal justice leaders and decision-makers as they
evaluate potential impacts of AI solutions on the community and other stakeholders.

A Design Thinking Approach to Identifying Use Cases

We advocate that all members of the criminal justice system take a design thinking approach to AI. Design thinking
is an approach to innovation that emphasizes deep understanding of the problem, its context, and constraints before
determining the best solution.6 When considering if and how AI might benefit prosecutors, a design thinking approach
begins with the question, “What is the problem to be solved or the job to be done?” For example, implementing
fair and just plea-bargaining strategies was identified as a challenge to improving case management and outcomes
in criminal cases (see Figure 2). To address this challenge, a design thinking approach would define the goals of plea
bargaining, the tools currently in use, and the challenges to achieving those goals. This approach would first identify what
the needs are for implementing fair, just plea-bargaining strategies and then assess the viability of an array of potential
solutions, including AI, to address those needs.

5. A full review of the ethical implications that criminal justice leaders should consider before implementing AI is beyond the scope of this brief. For a more complete review of ethical AI development, see
IEEE’s Ethically Aligned Design: A Vision for Prioritizing Human Well-being with Autonomous and Intelligent Systems, First Edition, and the AI Now Institute, an interdisciplinary research center focused on
understanding the social implications of AI. Both are included in the references at the end of this document.
6. Linke, R. (2017) Design thinking, explained. Retrieved from https://mitsloan.mit.edu/ideas-made-to-matter/design-thinking-explained

Artificial Intelligence Applications

3 for Criminal Courts
Most jobs that benefit from AI fall into one of two categories: performing tasks or making decisions. Only once the
context of the problem is well defined should prosecutors move on to the second question, “What is the best solution for
this problem?” In the case of AI, this question can be reframed as, “Is AI the best solution? If so, what level of automation
is appropriate?” Figure 3 provides examples of applications that help the criminal justice system perform tasks or make
decisions with differing levels of AI involvement.

Figure 3: Design thinking can help identify AI use cases by considering the level of AI involvement in the job that needs to be done.

A deeper discussion of key terms, considerations for success, and design thinking can be found in the first brief of this
series.

“The world is increasingly complex and full of data, and we have reached a point where there aren’t
enough humans to sort through the data and make the complex simple. AI provides us with the critical
capabilities needed to handle this big, complex world and to achieve an intelligence advantage.”

—Dr. Stacey Dixon, Director, Intelligence Advanced Research Projects Activity15

Artificial Intelligence Applications

4 for Criminal Courts
Opportunities and Examples of AI Addressing Prosecutor Needs
In 2019, RTI International—together with the RAND Corporation—published a report outlining the priority challenges
and needs of prosecutors and district attorneys.7 In addition, NIJ published strategic research priorities for the court
system in April 2020.8 The needs outlined in these documents serve as an excellent starting point for analyzing
opportunities for AI-enabled technology, as previously shown in Figure 2. Highlighting the examples that follow does not
mean that they (1) will be successful or (2) have been implemented ethically or cost effectively. Rather, these examples are
being highlighting to show new approaches that are being developed and tested. Awareness of emerging technologies
and products, as well as vendors, might enable future monitoring and adoption by the criminal court community.9

Managing Staffing and Resources

Although many of the challenges and needs related to staffing within the criminal courts are unique, the
difficulties of finding qualified talent are not. Other industries are currently using AI to increase the efficiency
of finding, validating, and hiring qualified candidates.10, 11 These staffing applications raise important concerns
about AI introducing negative bias in hiring, but some people argue that AI can eliminate bias in the hiring
process.12 Others point to AI’s ability to increase diversity in certain professions,13 a need that has been
identified through the NIJ’s Priority Criminal Justice Needs Initiative as relevant to the prosecutorial field.

Beyond hiring, resource constraints and the demands of the profession are two possible causes of high
turnover. AI may be able to increase efficiency of certain mundane or repetitive prosecutorial operations,
freeing up needed resources for higher-value activities. Many corporate law firms are using AI to automate
these more mundane tasks. For example, corporate counsel at JPMorgan Chase automated parts of their
contract review tasks and saved more than 360,000 hours of lawyers’ time.14

7. Lawrence, D. S., Gourdet, C., Banks, D., Planty, M. G., Woods, D., & Jackson, B. A. (2019). Prosecutor priorities, challenges, and solutions. Santa Monica, CA: RAND Corporation. Retrieved from https://www.
rand.org/pubs/research_reports/RR2892.html
8. National Institute of Justice. (2020). Courts Strategic Research Plan 2020–2024. (NCJ Number: 254684). Washington, DC: U.S. Department of Justice. Retrieved from https://nij.ojp.gov/library/
publications/courts-strategic-research-plan-2020-2024
9. Example products and vendor technologies are provided to serve as illustrative examples only. The Criminal Justice Testing and Evaluation Consortium does not endorse any specific product or vendor.
Mentions of companies and/or products do not represent approval or endorsement by the National Institute of Justice.
10. Heilweil, R. (2019). Artificial intelligence will help determine if you get your next job. Vox. Retrieved from https://www.vox.com/recode/2019/12/12/20993665/artificial-intelligence-ai-job-screen

11. Kasperkevic, J. (2019). Can AI make hiring fairer and more efficient? Marketplace. Retrieved from https://www.marketplace.org/2019/08/14/can-ai-make-hiring-fairer-and-more-efficient/

12. Polli, L. (2019). Using AI to eliminate bias from hiring. Harvard Business Review. Retrieved from https://hbr.org/2019/10/using-ai-to-eliminate-bias-from-hiring

13. Fisher, A. (2019) A.I. for hire: 4 ways algorithms can boost diversity in hiring. Fortune. Retrieved from https://fortune.com/2019/06/01/ai-artificial-intelligence-diversity-hiring/

14. Son, H. (2017). JPMorgan software does in seconds what took lawyers 360,000 hours. Bloomberg. Retrieved from https://www.bloomberg.com/news/articles/2017-02-28/jpmorgan-marshals-an-army-
of-developers-to-automate-high-finance
15. “AI for American Innovation.” (n.d.). https://www.whitehouse.gov/ai/ai-american-innovation/

Artificial Intelligence Applications

5 for Criminal Courts
 Improving Court Operations
Court operations include docket management, scheduling, security and cyber-security, facilities
management, evidence management, witness and juror management, and more. AI and other digital
technologies have the potential to improve court operations in numerous ways. Robotic process automation
(RPA) can be used to automate routine tasks or improve operational outcomes. Software vendors like
eCourtDate are attempting to improve operational efficiency and user-convenience by providing SMS
based reminders about court dates, payment notifications, and victim notifications.16 AI powered language
translation may help expand the availability of interpreter services in courts settings. In March 2020, Google
Translate released its first publicly available real-time transcription feature for Android users. Google’s mobile
app records speech in one language and transcribes translated text in real time.17

While AI adoption for improved court operations is still nascent within the US, other countries are exploring
digitally-enabled courts. In China, some court proceedings have even gone fully virtual, with a “’mobile court’
offered on popular social media platform WeChat that has already handled more than 3 million legal cases
or other judicial procedures since its launch in March 2019, according to the Supreme People’s Court…In a
demonstration, authorities showed how the Hangzhou Internet Court operates, featuring an online interface
in which litigants appear by video chat while an AI judge—complete with on-screen avatar—prompts them
to present their cases…Cases that are handled at the Hangzhou court include online trade disputes, copyright
cases and e-commerce product liability claims.”18

Processing and Managing Digital Information

The management and processing of digital information is where AI may have the most to offer the court
system in the near term. As the amount of digital information grows—driven by social media, body worn
cameras, and other digital devices—the sheer volume of digital evidence is growing rapidly. Since the 2012
Monique Da Silva Moore, et al. v. Publicis Groupe decision that allows for the use of technology-assisted
review,19 e-discovery technologies have been created to help attorneys sift through vast amounts of
electronic data to find data that may be relevant to a particular case. AI is improving these e-discovery tools
and may be able to help attorneys manage the increasing volume of case information.

Two mundane and time-consuming tasks that reviewers engage in are video redaction and audio
transcription. Advances in machine vision and natural language processing have improved software that
automates video redaction or transcribes audio files. These tools have expedited the law enforcement
reporting process, speeding up the review for criminal court cases. (See the second brief in this series for
additional details.) Although such applications still require human review prior to public release, these AI-
enabled solutions will likely continue to improve as AI technologies continue to evolve.

Data management and data quality form the foundation of successful AI implementations. AI should not be
seen as a tool to enable data management. Instead, good data management practices should be viewed as
prerequisites to creating internal systems that use AI. Given the data challenges outlined in the NIJ’s priority
needs assessment—along with the variability in technology infrastructure that exists in the court system at
large20— organizational data issues are likely to be a big barrier to AI adoption in many settings.

16. eCourtDate. (2020). Everyone Needs a Reminder. Court date reminders, victim notifications, case alerts, payment notices. Washington, DC: eCourtDate. Retrieved from https://ecourtdate.com/
17. Statt, N. (2020, March 17). Google Translate’s real time transcription feature is out now for Android. The Verge. Retrieved from https://www.theverge.com/2020/3/17/21182640/google-translate-
transcription-android-feature-real-time-ai
18. No Author. (2019, January 16, 2020). In brave new world of China’s digital courts, judges are AI and verdicts come via chat app. Tokyo, Japan: The Japan Times. Retrieved from https://www.japantimes.
co.jp/news/2019/12/07/asia-pacific/crime-legal-asia-pacific/ai-judges-verdicts-via-chat-app-brave-new-world-chinas-digital-courts/#.Xr2uCBNKgUs
19. Verga, M. (2018, July 31). In the beginning was da Silva Moore (Technology-assisted Review Series, 2). XACT Data Discovery. Retrieved from https://www.xactdatadiscovery.com/articles/in-the-
beginning-was-da-silva-moore/
20. Jackson, B. A., Banks, D., Hollywood, J. S., Woods, D., Royal, A., Woodson, P. W., & Johnson, N. J. (2016). Fostering innovation in the U.S. court system: Identifying high-priority technology and other needs
for improving court operations and outcomes. Santa Monica, CA: RAND Corporation. Retrieved from https://www.rand.org/pubs/research_reports/RR1255.html

Artificial Intelligence Applications

6 for Criminal Courts
 Improving Case Management and Outcomes
Case management includes evidence management, analyzing case law, witness and community engagement,
multimedia, plea bargaining, risk assessments, pretrial legislation, sentencing guidelines, and more. One
way that AI is improving case management is increasing the efficiency of legal research. As natural language
processing improves, software tools will become increasingly sophisticated in their ability to assist with legal
research. Major legal support software vendors are already advertising that their products are AI-enabled,
increasing the efficiency with which prosecutors can conduct legal research or find relevant case law.21, 22

Additionally, AI may help inform litigation strategies. Decisions about how to proceed after charges are filed
are some of the most important and consequential determinations that a prosecutor makes. Such decisions
often rely on attorneys weighing available evidence alongside their own experience. Although still a nascent
application, AI may one day be used to predict trial outcomes based on available evidence and inform
prosecutorial decision-making and strategy development. For example, some academics have used machine
learning and predictive analytics to predict outcomes in court decisions, including decisions for the European
Court of Human Rights23 and the Supreme Court of the United States.24

Lastly, risk assessment tools are becoming more prevalent in courtrooms across the US.25 Proponents of these
systems highlight the potential to reduce human bias in decision-making.26 Detractors point to instances in
which such systems have perpetuated the systemic bias embedded in data used to create these systems.27
An in-depth discussion of risk assessment tools is beyond the scope of this brief. However, prosecutors and
public defenders should continue to have ongoing conversations that will ultimately determine if and how
these systems are used. For further reading about risk assessment tools, see Handbook of Recidivism Risk/
Needs Assessment Tools, First Edition,28 and Report on Algorithmic Risk Assessment Tools in the U.S. Criminal Justice
System.29

It is widely accepted that human decision-making is imperfect. Sometimes this is due to incomplete
information or lack of critical insights. Other times, this is due to human bias or errors in judgment. Anyone
deploying AI systems to make decisions or recommendations must wrestle with the question: Is perfect
decision-making a prerequisite to deployment? Or is improvement over existing alternatives sufficient?

21. Announcing Westlaw Edge. (n.d.). Retrieved from https://legal.thomsonreuters.com/en/insights/articles/announcing-westlaw-edge

22. 2020 legal analytics study: Bringing value into focus. (n.d.). Retrieved from https://www.lexisnexis.com/en-us/products/lexis-analytics.page
23. Medvedeva, M., Vols, M. & Wieling, M. (2019). Using machine learning to predict decisions of the European Court of Human Rights. Artificial Intelligence and Law. https://doi.org/10.1007/s10506-019-
09255-y
24. Katz, D. M., Bommarito, M. J., II, Blackman, J. (2017). A general approach for predicting the behavior of the Supreme Court of the United States. PLOS One. https://doi.org/10.1371/journal.pone.0174698

25. Algorithms in the criminal justice system: Pre-trial risk assessment tool. (n.d.). Epic. Retrieved from https://epic.org/algorithmic-transparency/crim-justice/

26. Watney, C. (2017). It’s time for our justice system to embrace artificial intelligence. Brookings. Retrieved from https://www.brookings.edu/blog/techtank/2017/07/20/its-time-for-our-justice-system-to-
embrace-artificial-intelligence/
27. Hao, K. (2019). AI is sending people to jail—and getting it wrong. MIT Technology Review. Retrieved from https://www.technologyreview.com/s/612775/algorithms-criminal-justice-ai/

28. Singh, J. P., Kroner, D. G., Wormith, J. S., Desmarais, S. L., & Hamilton, Z. (Eds.). (2017). Handbook of recidivism risk/needs assessment tool. Hoboken, NJ: Wiley.

29. Partnership on AI. (n.d.). Report on algorithmic risk assessment tools in the U.S. criminal justice system. Retrieved from https://www.partnershiponai.org/report-on-machine-learning-in-risk-
assessment-tools-in-the-u-s-criminal-justice-system/

Artificial Intelligence Applications

7 for Criminal Courts
 Maintaining Accountability
Requirements related to Brady and Giglio, data management procedures, and pressures from conviction
integrity units create needs related to maintaining accountability. Some district attorneys are finding
creative ways to use AI to try to improve accountability. The San Francisco District Attorney is using AI
redaction software in an attempt to eliminate bias in decisions about who to prosecute. Prosecutors use an
AI-enabled “bias mitigation” tool developed at Stanford to “automatically redact information from police
reports that could identify a suspect’s race. [The tool] is designed to be a way to keep prosecutors from being
influenced by racial bias when deciding whether and how someone gets charged with a crime…The tool
will not only strip out descriptions of race, but also descriptors like eye color and hair color, according to the
district attorney’s office. The names of people, locations, and neighborhoods that might all consciously or
unconsciously tip off a prosecutor that a suspect is of a certain racial background are also removed.”30

Addressing the needs of prosecutors could contribute to improving the efficiency, legitimacy, and
administration of justice within prosecutors’ offices and the criminal justice system, as well as in the eyes of
the victims and the community.31

Creating Partnerships and Collaboration

Criminal courts leaders can benefit from partnering and collaborating with others on the development and
implementation of AI tools. Those looking to deploy AI may partner with data scientists, researchers, and
other partners to support AI adoption in criminal court functions, as appropriate. These partners can help
translate information from one system to another, code information to train AI algorithms, work to examine AI
implications, and help determine whether AI systems can truly support efficiencies while also maintaining fair
and individualized justice. Partnerships may also help create value from the massive amount of information
that prosecutors and courts maintain. Partners and collaborators can help answer some persistent questions
that can promote public transparency, such as determining if there is consistent decision-making within
and across jurisdictions or understanding the number and characteristics of cases that are disposed via plea
bargains versus those that go to trial.32

30. Hollister, S. (2019). San Francisco says it will use AI to reduce bias when charging people with crimes. The Verge. Retrieved from https://www.theverge.com/2019/6/12/18663093/ai-sf-district-attorney-
police-bias-race-charge-crime
31. Lawrence, D. S., Gourdet, C., Banks, D., Planty, M. G., Woods, D., & Jackson, B. A. (2019). Prosecutor priorities, challenges, and solutions. Santa Monica, CA: RAND Corporation. Retrieved from https://www.
rand.org/pubs/research_reports/RR2892.html
32. National Research Council 2001. What’s Changing in Prosecution?: Report of a Workshop. Washington, DC: The National Academies Press. https://doi.org/10.17226/10114.

Artificial Intelligence Applications

8 for Criminal Courts
Beyond needs identified through the Priority Criminal Justice Needs Initiative, there are other
noteworthy examples of AI being applied within the court system.
Assisting individual citizens and pro-se litigants
A 2016 report highlighted the increasing number of pro-se litigants who may need help navigating the legal system.33
AI-enabled legal recommendation services could help these self-represented litigants navigate the legal system.
Systems such as virtual legal chatbots and legal recommendation software may help create more equitable access
to legal services—especially with instances in which knowledge gaps exist. Some organizations are going one step
further and assisting citizens that don’t even know they are eligible for assistance. Code for America is developing an AI
application for expungement. They are working with the internal databases of California courts to automatically identify
expungement-eligible records, which eliminates the need for individuals to apply.34 A United Kingdom-based company
created a legal assistance app called DoNotPay that helped overturn 160,000 parking tickets in London and New York
City.35

Detecting digital evidence tampering

AI has enabled new forms of video manipulation, which raises issues during litigation. AI-enabled “deepfakes” or other
doctored videos and photos can be difficult to detect. AI researchers, including those at Google, are working to detect
evidence of tampering or doctoring of digital images.36 Such technology may help authenticate digital evidence as
“undoctored.” However, some experts worry that detection technology will never be able to keep pace with deepfake
technology.37

Recommending asset split in a divorce

The Federal Court of Australia has created a machine learning proof of concept that helps parties in divorce cases divide
assets and liabilities.38 The system is an example of AI assisting, rather than replacing, human decision-making because
the system’s recommendations are nonbinding and are simply a tool used during negotiations.

Informing judicial decisions

In addition to using AI in risk assessment and other prediction tools, AI can also be used to identify what data are likely
most useful to inform a judicial decision, pull that relevant data from identified sources, and aggregate the information
into a “snapshot” or dashboard to inform judicial decision-making. This approach differs from more traditional data
dashboards that display information in the same way for each case. A juvenile court judge in Ohio used AI technology to
analyze data for juvenile offenders and to create a multipage summary for each offender.39 In Buenos Aires, the AI startup
Prometea is used to draft legal documents and suggest rulings to judges. The country of Estonia has even more ambitious
plans and is attempting to create an “AI Judge” to fully automate judicial decision-making for certain small claims cases.40

33. Jackson, B. A., Banks, D., Hollywood, J. S., Woods, D., Royal, A., Woodson, P. W., & Johnson, N. J. (2016). Fostering innovation in the U.S. court system: Identifying high-priority technology and other needs
for improving court operations and outcomes. Santa Monica, CA: RAND Corporation. Retrieved from https://www.rand.org/pubs/research_reports/RR1255.html
34. Rosenberg, T. (2016, June 1). Legal aid with a digital twist. The New York Times. Retrieved from https://www.nytimes.com/2016/06/01/opinion/legal-aid-with-a-digital-twist.html

35. Niiler, E. (2019, March 25). Can AI be a fair judge in court? Estonia thinks so. Wired. Retrieved from https://www.wired.com/story/can-ai-be-fair-judge-court-estonia-thinks-so/

36. Metz, C. (2019, November 24). Internet companies prepare to fight the ‘deepfake’ future. The New York Times. Retrieved from https://www.nytimes.com/2019/11/24/technology/tech-companies-
deepfakes.html
37. Engler, A. (2019, November 14). Fighting deepfakes when detection fails. Brookings. Retrieved from https://www.brookings.edu/research/fighting-deepfakes-when-detection-fails/

38. Crozier, R. (2019, May 23). Fed court turns to AI to predict asset split after relationship breakdown. IT News. Retrieved from https://www.itnews.com.au/news/fed-court-turns-to-ai-to-predict-asset-split-
after-relationship-breakdown-525587
39. Goodman, C. C. (2019). AI/Esq.: Impacts of artificial intelligence in lawyer-client relationships. Oklahoma Law Review, 72(1). Retrieved from https://digitalcommons.law.ou.edu/cgi/viewcontent.
cgi?article=1380&context=olr
40. Gillespie, P. (2018, October 26). This AI startup generates legal papers without lawyers, and suggests a ruling. Bloomberg Businessweek. Retrieved from https://www.bloomberg.com/news/
articles/2018-10-26/this-ai-startup-generates-legal-papers-without-lawyers-and-suggests-a-ruling

Artificial Intelligence Applications

9 for Criminal Courts
Five AI Considerations for Criminal Courts
1. The use of AI is not yet widespread in criminal courts. However, AI implementations are increasing, and new
applications are emerging. AI is likely to transform many aspects of the legal profession in the years to come.
2. A design thinking approach can help prosecutors and other officers of the court identify AI use cases. First, assess
the jobs to be done and identify problems of practice. Then, look at various technologies, including AI, to identify the
best solutions. Court system leaders can mitigate AI risks by carefully considering the impacts of bias in AI systems
that provide recommendations or predictions.
3. Prosecutors and other officers of the court may decide to pursue lower-risk applications in the near term as society
wrestles with the larger implications of algorithmic decision-making; prosecutors may take this approach because
of concerns about fairness and transparency. Examples include improvements to operational efficiency rather than
higher-risk applications that make recommendations or decisions.
4. Implementation may require substantial process or procedural changes, regardless of the technical readiness of AI
systems. Deploying AI-enabled technology will succeed only if implementers are given sufficient time and resources.
5. Early adopters in other law disciplines (e.g., corporate law) and even other professions (e.g., finance or health care)
can serve as bellwethers for the costs, benefits, and risks of implementing AI systems.

Future Outlook
AI is here to stay, and advances in technical capabilities will continue. The criminal justice community faces shrinking budgets
and a growing sense of mistrust from the community. With these things in mind—and considering ethical appropriateness,
technical feasibility, and the operational limitations—AI provides important opportunities to improve the criminal justice
system. Opportunities to implement AI tools should be met with a clear understanding of the data requirement and use a
design thinking approach to evaluating potential use cases. This series of briefs aims to inform decision makers about what is
already happening in the criminal justice ecosystem and what is required to utilize emerging AI technologies in a thoughtful,
informed, and unbiased way.

The NIJ continues to support a portfolio of AI research projects in areas such as public safety video and image analysis, DNA
analysis, gunshot detection, and crime forecasting.41 Looking ahead to the future, different countries and states are likely to
adopt AI technologies for criminal justice applications at different rates, which presents an opportunity for learning through
collaboration. Improving criminal justice outcomes through the use of AI-enabled technologies will require intentional
investment, careful consideration, and sustained efforts from criminal justice decision makers. If designed and implemented
well, AI-enabled tools have the potential to improve efficiency, reduce costs, and expand capabilities across many criminal
justice use cases.

41. Christopher Rigano, “Using Artificial Intelligence to Address Criminal Justice Needs,” October 8, 2018, nij.ojp.gov: https://nij.ojp.gov/topics/articles/using-artificial-intelligence-address-criminal-justice-
needs

Artificial Intelligence Applications

10 for Criminal Courts
Additional References Published: August 2020

Berkman Klein Center for Internet & Society at Harvard University. (2019, October 15). AI: Algorithms and justice. Retrieved from https://
cyber.harvard.edu/projects/ai-algorithms-and-justice More Information
Bird & Bird LLP. (2019, May 9). AI in the courtroom. Digital Business. Retrieved from https://digitalbusiness.law/2019/05/ai-in-the-
courtroom/#page=1 Steven Schuetz
Senior Science Advisor
Brundage, M., Avin, S., Clark, J., Toner, H., Eckersley, P., Garfinkel, B., …, & Amodei, D. (2018). The malicious use of artificial intelligence: National Institute of Justice
Forecasting, prevention, and mitigation. Retrieved from https://img1.wsimg.com/blobby/go/3d82daa4-97fe-4096-9c6b-376b92c619de/
downloads/MaliciousUseofAI.pdf?ver=1553030594217
U.S. Department of Justice
Steven.Schuetz@usdoj.gov
Cheung, K. C. (2020, January 9). Top 10 applications of AI in law. Algorithm-XLab. Retrieved from https://algorithmxlab.com/blog/top-10- Tel +1-202-514-7663
applications-artificial-intelligence-in-law/

Donahue, L. (2018, January 3). A primer on using artificial intelligence in the legal profession. Jolt Digest. Retrieved from https://jolt.law. Jeri D. Ropero-Miller, PhD, F-ABFT
harvard.edu/digest/a-primer-on-using-artificial-intelligence-in-the-legal-profession Project Director, CJTEC
European Commission for the Efficiency of Justice. (n.d.) Possible use of AI to support the work of courts and legal professionals. Council of RTI International
Europe. Retrieved from https://www.coe.int/en/web/cepej/tools-for-courts-and-judicial-professionals-for-the-practical-implementation- jerimiller@rti.org
of-ai Tel +1-919-485-5685
Fagella, D. (2020, March 14). AI in law and legal practice: A comprehensive view of 35 current applications. Emerj. Retrieved from https://
emerj.com/ai-sector-overviews/ai-in-law-legal-practice-current-applications/ Duren Banks
Goodman, C. C. (2019). AI/Esq.: Impacts of artificial intelligence in lawyer-client relationships. Oklahoma Law Review, 72(1). Retrieved from Vice President
https://digitalcommons.law.ou.edu/cgi/viewcontent.cgi?article=1380&context=olr Division for Applied Justice Research
RTI International
Lavinder, J. (2019, October 9). The artificial intelligence revolution in legal services. Law. Retrieved from https://www.law.com/
durenbanks@rti.org
corpcounsel/2019/10/09/the-artificial-intelligence-revolution-in-legal-services/
Tel +1-919-541-8026
Legal tech market map: 50 startups disrupting the legal industry. (2016, July 13). CB Insights. Retrieved from https://www.cbinsights.com/
research/legal-tech-market-map-company-list/
James Redden
Lopez, I. (2016, July 25). But what about lawyers? A Q&A with Richard Susskind on AI in law. Law. Retrieved from https://www.law.com/ Innovation Advisor
legaltechnews/almID/1202763509782/
RTI International
Manyika, J., Chui, M., Miremadi, M., Bughin, J., George, K., Willmott, P., & Dewhurst, M. (2017). A future that works: Automation, jredden@rti.org
employment, and productivity. McKinsey and Co. Retrieved from https://www.mckinsey.com/~/media/McKinsey/Featured%20Insights/ Tel +1-919-248-4222
Digital%20Disruption/Harnessing%20automation%20for%20a%20future%20that%20works/MGI-A-future-that-works_Full-report.ashx

Manyika, J., Chui, M., Miremadi, M., Bughin, J., George, K., Willmott, P., & Dewhurst, M. (2017). Harnessing automation for a future that
works. McKinsey and Co. Retrieved from https://www.mckinsey.com/featured-insights/digital-disruption/harnessing-automation-for-a-
future-that-works Suggested Citation
Maras, M.-H. and Alexandrou, A. (2019) Determining authenticity of video evidence in the age of artificial intelligence and in the wake of
Deepfake videos, The International Journal of Evidence & Proof, 23(3), pp. 255–262. doi: 10.1177/1365712718807226. Redden, J., Banks, D., & Criminal Justice Testing and
Evaluation Consortium. (2020). Artificial Intelligence
Martin, M. (Host). (2019, June 15). San Francisco DA looks to AI to remove potential prosecution bias. All Things Considered. Retrieved from
https://www.npr.org/2019/06/15/733081706/san-francisco-da-looks-to-ai-to-remove-potential-prosecution-bias Applications for Criminal Courts. U.S. Department of
Justice, National Institute of Justice, Office of Justice
Polonski, V. (2018, November 19). AI is convicting criminals and determining jail time, but is it fair? World Economic Forum. Retrieved from Programs. http://cjtec.org/
https://www.weforum.org/agenda/2018/11/algorithms-court-criminals-jail-time-fair/

Rawlinson, P. (2018, March 29). Will lawyers become extinct in the age of automation? World Economic Forum. Retrieved from https://www.
weforum.org/agenda/2018/03/will-lawyers-become-extinct-in-the-age-of-automation/

Re, R. M., & Solow-Niederman, A. (2019). Developing artificially intelligent justice. Stanford Technology Law Review, 22(2). Retrieved from
https://law.stanford.edu/publications/developing-artificially-intelligent-justice-stanford-technology-law-review/ This publication was made possible by Award Number
Schieneman, K., & Gricks, T. C., III. (2013). The implications of rule 26(g) on the use of technology-assisted review. The Federal Courts 2018-75-CX-K003, awarded by the National Institute
Law Review, 7(1). Retrieved from https://www.schnader.com/files/Publication/df9ff801-13aa-45c5-88e8-4afb6b5e032a/Presentation/ of Justice, Office of Justice Programs, U.S. Department
PublicationAttachment/46e54cf3-47c0-4033-87ff-73e1a59ed2ea/Rule%2026Gricks.pdf of Justice. The opinions, findings, and conclusions or
Thomson Reuters. (2017). Ready or not: Artificial intelligence and corporate legal departments (Legal Department 2025). Thomson Reuters.
recommendations expressed in this publication are
Retrieved from https://static.legalsolutions.thomsonreuters.com/static/pdf/S045344_final.pdf those of the author(s) and do not necessarily reflect
those of the Department of Justice.
World Economic Forum. (2018). The future of jobs report (Insight Report). Retrieved from http://www3.weforum.org/docs/WEF_Future_
of_Jobs_2018.pdf

Wu, J. (2019, August 5). AI goes to court: The growing landscape of AI for access to justice. Medium. Retrieved from https://medium.com/ http://cjtec.org/
legal-design-and-innovation/ai-goes-to-court-the-growing-landscape-of-ai-for-access-to-justice-3f58aca4306f

Artificial Intelligence Applications

11 for Criminal Courts