RaeMurphy - EST 110 by Rae Murphy

Personalized Overview

raemurphy — 2025-03-17 02:46:06 UTC

For this, I chose to do how AI is currently being used in the policing and justice system. I chose this field specifically because I feel like there’s a lot that I could be used in and many that are going to improve their effectiveness. I also think that the ways it’s used are interesting in a way that many other fields of work don’t use AI for. I’ll talk about multiple uses of AI in both policing and justice systems, how they intercept, and how they help improve functions. I’ll talk about what these ai usages mean for both civilians and people who work in the system as well as what we could possibly see it doing in the future. What does this technology mean for privacy, fairness, and ethics, and what are the patterns we’re seeing in it currently? I personally thought that the system was interesting because I feel like it’s very common and because of the crime database and how frequent crime and cases are. Given the ongoing expansion of AI and the increasing volume of data generated from crime reports and case files this technology is not only rapidly expanding, but also highly relevant. It’s important to think about how it could impact people’s lives, jobs and how cases may be handled in the future, and what new precautions may have to be accounted for.

Example 1: Law Enforcement - Facial Recognition

raemurphy — 2025-03-17 02:49:11 UTC

In law-enforcement, facial recognition is used to do many things. It can be used to search through billions of photos to identify suspects of crimes that are currently unknown, or even be used to confirm if a suspect that is in custody is truly a face seen on cctv. It uses machine learning and deep learning algorithms to recognize unique faces and patterns in provided evidence. This helps speed up the process of finding potentially unknown suspects, or ruling out already known suspects. However, if unregulated or without policy, any mistakes made by the technology may lead to wrongful arrests or possible cases being thrown out.

https://youtu.be/YX8BzK_LU0E?si=20igS4KSELjnO5wW

Video ^^

Example 2: Law Enforcement - License Plate Recognition

raemurphy — 2025-03-17 02:51:14 UTC

In law enforcement, license plate recognition AI is used to capture images of passing vehicles, including license plates. Algorithms analyze these images and the tax plate numbers along with other car details. These systems use ALPR (Automatic License Plate Reader) technology to better conduct investigations within the area of a crime scene, efficiently and effectively. This is useful in a multitude of ways, it can be added to the database and be compared against multiple different previously taken numbers and vehicles. They can also alert cops of a potential suspect vehicle and its movements. Of course, mistakes can happen and an Ai reader can be wrong/make an incorrect conclusion about a vehicle, however these are small bumps in the road that are slowly being smoothed out.

https://youtu.be/COBClFEQQJk?si=nRWTJCeCxNqotbpP

Video^^

Example 3: Police Work - Gunshot Detection

raemurphy — 2025-03-17 02:53:04 UTC

Gunshot detection systems are used in policing to identify, locate, and respond to gunfire incidents in real time. The systems use AI powered extensors and machine learning algorithms and help law enforcement reach a potential violent crime scene in improved response times. It has many useful features like location tracking and sound analysis to determine whether a possible gunshot actually did go off and where it happens, and if police are needed. Because of the way it works, sometimes false positives can come through. For example, sometimes an AI can identify fireworks backfires or construction noises gunfire meaning that police responses may be necessary in the aftermath.

https://youtu.be/QNJYWwlb7xY?si=LuD5w5zGQheOXbTC

Video ^^

Future Trends and Ethical Concerns

raemurphy — 2025-03-17 02:53:38 UTC

With the way things go, AI is constantly improving, not only in this field, but in all aspects of life. But of course that applies to law enforcement, meaning that AI will only get better and better and hopefully more accurate and. However, many things can pop up and hold us back from that potential future. Bias discrimination can be a big issue as systems were for historical data, which often reflects existing bias in policing. This can be tied to race, socioeconomic groups, or even gender/disabilities. I can hopefully be trained on data that is more diverse and representative, and regulate its own biases in order to more accurately reflect the data being given. It’s also an issue to do with privacy, as things like facial recognition, predictive policing, and gunshot protection, raise privacy concerns. A lot of AI and surveillance has led to concern with civilians. It’d be better to limit surveillance to cases with probable cause rather than just broad and general monitoring in order to make civilians feel more comfortable and respected within their own environment, and also to prevent potential misguided assumptions within investigations. The best way to move forward with AI in law enforcement is with accountability and transparency, making sure that everyone knows their rights and that all proper regulations are put in place in order for this technology to be used in the best way it can be and not abused. It should be a support in law enforcement, not replacing humane judgment which we have within our practices. It should be ethical, transparent, and fair, ensuring that it serves justice rather than just trying to put someone behind bars.

Societal Impact

raemurphy — 2025-03-17 02:54:27 UTC

Ai was made to be convenient for workers, assisting them in menial tasks so that they could focus on bigger and better things. But of course, as it gets better and better, the question is, will it start to outperform humans?

A few concerns when it comes to jobs is that crime analysts may become more absolute, as well as automated surveillance leading to less of a need for traffic officers. A lot of these fundamental positions could be handed off to an AI but are still developing and still making mistakes that human minds wouldn’t, which may be hasty and lead to rougher roads down the line. Yeah I should complement not replace human officers, while automation may reduce the need for certain aspects of jobs, at the end of the day law enforcement is upon humans that should be done by humans and not have everything run by a machine. On top of that privacy is of course, becoming a huge concern in general, but especially because of law enforcement due to the idea of rights being breached. Monitoring makes people uncomfortable, as well as makes them wary when most people won’t commit crimes in their life. Data collection can be used against people and can be Used wrongfully, which is a major concern for many. Ai should be regulated in order to prevent such abuse, because even though crime prevention is important, sacrificing privacy for security leads to potential corruption.

Overview Video

raemurphy — 2025-03-17 02:58:31 UTC

https://drive.google.com/file/d/1PzhQkvG9U-oWdcQ5PugKXJ-7nOnTfnJ-/view?usp=drivesdk

(Video embed feature just didn’t work for me 🙁)

Reflection

raemurphy — 2025-03-17 02:58:50 UTC

Personally, I think AI is actually a pretty useful tool in law-enforcement, it does a lot of things that would take many people a lot of time. Combing through data saves those people time and gives them more opportunities to go look for important pieces of evidence or prepare defense and jury or cases that could be very important. I never really thought that AI was so prevalent in law-enforcement, despite knowing that it was at least their part. But now that I know that it’s slowly creeping into every aspect of law-enforcement and that seems a bit scary, I also feel that if driven in the correct direction, it could be very useful.

References

raemurphy — 2025-03-17 02:59:41 UTC

https://www.ojp.gov/pdffiles1/nij/252038.pdf

https://www.gao.gov/products/gao-24-107372

https://www.utility.com/blog/revolutionizing-license-plate-recognition-with-pulsars-ai-powered-solutions/

https://naacp.org/resources/artificial-intelligence-predictive-policing-issue-brief#:~:text=Bias%20and%20Discrimination%3A%20AI%20models,policing%20and%20resources%20are%20made.

https://www.policingproject.org/ai-explained-articles/2024/9/6/how-policing-agencies-use-

ai#:~:text=Gunshot%20detection%20systems%20use%20a,evidence%20such%20as%20shell%20casings.