Artificial Intelligence (AI) and Policing
Artificial intelligence technology is making its mark on every aspect of our personal and professional lives. The police are no exception. For several years, police officers have been using software for facial recognition, crowd monitoring and crime prevention.
A few decades ago, AI technology was little more than the subject of science fiction movies and novels. Today, this type of technology is entering all aspects of our lives, including law enforcement. As for the police, robots help in monitoring and guarding low-security areas like malls and high-security and risky areas like power plants, construction sites, etc.
AI technology uses algorithms to analyze huge amounts of data in less time. By studying human behavior, software also develops the ability to mimic and possibly predict future actions. As the capabilities and accuracy of the technology grow, AI is likely to become even more widespread in law enforcement.
Experts believe that smart technologies like AI can help reducing urban crime by up to 40%. Artificial intelligence can also reduce emergency response rates by 20 to 35%. Despite these obvious advantages of AI, when it comes to protecting citizens, the technology is not without controversy. Most of these concerns are related to predictive policing and surveillanceaccording to consultants Deloitte.
Facial recognition technology is one of the most popular applications of AI technology. Facial recognition software allows police officers to identify individuals beyond doubt. They no longer have to manually check IDs in different databases. Besides recording an actual image, most of these software applications also collect biometric data. Biometric information allows for more accurate identification. There are some challenges with facial recognition technology, but they can be supplemented with biometric information to improve their accuracy.
Globally, law enforcement agencies use facial recognition technology to:
- Find wanted persons more easily
- Identify people included in images with less risk of false positives
- Identify injured or unconscious victims in traffic accidents
- Retrospectively confirm an individual’s identity and verify it against existing databases
Thanks to significant development over the past few years, facial recognition technology can now also be used live. Live Face Recognition (LFR) compares camera feeds with watch lists of known and wanted criminals, for example. Because it works in real time, LFR allows police forces to arrive on scene within minutes when the software detects a match.
There is definitely a very serious concern about bias in face recognition, and that stems from the biased datasets we feed it to learn from. As we improve these datasets and use different data to learn from these machines, they will only get better.
Also Read: Artificial Intelligence and Disinformation.
Beyond facial recognition
Establishing the identity of wanted persons has always been an important part of police work. AI can further improve this process, but the real power of this technology lies in crime prediction and prevention.
Artificial intelligence software can analyze unimaginable amounts of data, for example from CCTV feeds. In addition to looking for faces, the software also identifies trends, patterns of behavior and other correlations much faster than humans could. Technology far surpasses humans when it comes to the amount of data that needs to be analyzed.
While analytics forms the foundation of all AI applications, machine learning then allows software to draw human conclusions. Based on these results, AI can predict the future. The process may sound straightforward, but machine learning takes time and several iterations before the algorithm makes meaningful conclusions.
Human behavior is complex and often driven by a variety of motives. It is theoretically possible for software to learn and implement all of them in the future. Currently, however, AI is playing a supporting role in law enforcement and policing. Technology is not yet capable of taking over from human officers.
For example, based on its data analysis, AI software can identify behavioral patterns and make predictions about potential future crimes based on them. But predictive policing based solely on technology remains controversial. However, this type of policing may be the main style of policing in the future.
Reducing police paperwork
Police forces around the world complain about the amount of paperwork officers have to fill out after incidents they attend. Creating and updating case files keeps officers off the streets and can endanger the safety of citizens.
To say that case reports are reduced would also be problematic, as they are often the basis of successful criminal prosecution. AI can help by automatically capturing the necessary data, thereby minimizing the time employees spend on reporting. Employees may need to review and annotate the collected data, but they are likely to spend far less time than it would take to complete the entire process by hand.
Recording data through AI technology and its fact-checking afterwards not only reduces the time required. It also helps minimize the potential for human error or bias in the report.
Intelligent incident knowledge sharing
Police incident reports gathered dust in the archives, where various departments kept hard copies of crime reports and investigations. AI technology combined with collaboration software makes it easier to share information between departments and agencies.
Sharing information often means accessing different databases and comparing their contents. Done by a single employee or even a team of employees, it will take hours, if not days. AI, on the other hand, can easily cross-reference the content of multiple databases and share its conclusions.
Police forces are not only gaining access to more information. They also benefit from having an invaluable “team member” who digests vast amounts of data and draws human conclusions from it.
Intelligent knowledge sharing of this type benefits each of the police forces and law enforcement agencies involved.
The successful use of AI technology in policing is based on confidence and mutual trust. This trust must exist between different law enforcement agencies when it comes to sharing data. It is also required between the police and their community.
Robots and security
Robots work on improving, monitoring and security in low and high risk areas by patrolling malls, power grids…etc. These robots are used to reach areas inaccessible or unfavorable for human patrolling or surveillance.
Remote monitoring and inspection
Drones can provide critical remote monitoring and inspections performed without human intervention in the area being monitored or patrolled. The drone’s aerial capabilities allow it to inspect structures that are difficult to reach from the ground.
Researchers from the University of Maryland and the University of Zurich equipped a drone with event cameras and a sonar system so it could detect and avoid objects thrown at it. These drones can be used to intervene in high-risk environments without requiring police force to harm.
A robotic police force
The Huntington Park Police Department has introduced its newest recruit, a 400-pound robot known as HP RoboCop. Since then, he has been patrolling the Salt Lake Park in California and helping make arrests. Just imagine a group of these robots working in unison to spread a protective blanket around the city, it could seriously reduce crime in the area. In particular, this robot has helped the police department capture criminals with evidence in 6-8 hours!
Without a certain level of trust and acceptance, smart, innovative community policing cannot deliver on its promise. When citizens feel they are being watched without being able to feel safer, they will not see the benefit of AI technology for policing.
As the capabilities of AI technology grow and predictive policing becomes more of a reality, community trust must grow as well. AI will only reach its full potential in policing when trust and technology truly come together.