Unfortunately, crime is everywhere around us. Every city has a police department tasked with apprehending criminals and lowering crime rates. The police's primary goal is to keep crime under control. No one can deny it. Indeed, professional crime-fighting is widely regarded as the most effective policy approach because it embodies a strong commitment to this goal. To deal with crime in a smart way, smart cities necessitate smart policing.

The adoption of technology by the Indian police has been increasing at a rapid rate over the last few years. Blockchain, artificial intelligence (AI), advanced biometrics recognition systems, drones, body-worn cameras, cryptocurrency analytics, and cloud forensics are among the tools used by police around the globe.

The Uttar Pradesh Police, for example, uses an AI-enabled app created by the start-up Staqu. The software uses artificial intelligence to digitize and scan documents, as well as carry criminal records, aiding police forces on the ground with real-time information collection during inquiries, routine checks, and verifications, as well as at police checkpoints.

Fortunately, the app also includes a feature called Gang Recognition Technology, which aids police in not only detecting a suspect but also their associates who are involved in various districts and states. Punjab, Rajasthan, and Uttarakhand are among the police departments with which the start-up is collaborating. The importance of using technology is enormous: the start-up has assisted the police in the resolution of over 400 high-risk and complex incidents. 
 
The Concept of Smart Policing
Smart policing is a new approach of Indian policing that emphasizes crime prevention and encourages the strengthening of policing's evidence base. Smart policing focuses on using data and analytics efficiently, enhancing monitoring, success assessment, appraisal research, as well as increasing productivity and promoting creativity.

This introduction describes Smart Policing in contemporary and historical contexts, as well as several keys and evolving features in local Smart Policing locations. Specifically, the need to strengthen policing's evidence base, the emerging police-research collaborations in Smart Policing, the types of problems found and interventions taken, and possible Smart Policing concerns. 
 
Smart Policing in the Age of Artificial Intelligence
There are likely to be worrying questions for police officials who are currently in charge of India's rising smart cities such as Should we wait for technology to advance before making a bold decision? Are we prepared enough to handle anything that comes our way? Is it going to cost a lot of money? Given the buzz surrounding Artificial Intelligence (AI), it's reasonable to ask these questions, but it seems Artificial Intelligence (AI) is on the verge of being unstoppable, and there's no turning back, it's here to stay. Now it's only a matter of time before police forces all over the country begin to use them to make our communities safer. 
 
How Artificial Intelligence is modifying the Police
Governments have seen AI's benefits in a variety of fields, including banking, healthcare, insurance, and transportation. Governments are now adopting AI approaches in policing to tackle crimes and terror attacks in their territories, thanks to the rapid decline in the cost of data processing. The Andhra Pradesh Police Department has created an app to assist in the tracking of old criminals and suspects. This is intended to bridge the distance between where an offender lives and where he or she commits a crime.

Unless the head of the police station where the crime is committed reports the head where the offender lives, the head of the police station where the offense is committed may not be aware of the offender's activities. Since all records of history sheeters are digitized and their operations are updated daily, the state now has a clearer track record in combating crime. Certain data can be fed into facial recognition systems, and their movement in sensitive areas including airports can be identified. AI software is now being used by many police forces around the world to forecast crimes and detect suspicious people. 
 
The Importance of Intelligence for More Effective Policing
Since it is difficult for humans to deal with such vast and complex data, AI not only improves police performance but also offers significant input from data and aids in preventing crime and upholding law and order. As part of the Indian Digital Police initiative, we now have databases that exchange knowledge about crimes with various police forces, such as the Crime and Criminal Tracking Network and Systems.

Artificial intelligence (AI) software could help in detecting police brutality and avoiding escalation, which would contribute to a prison's already stressful climate. Guards and other duty staff who are otherwise functioning with repressed issues will increase their chances of finding support by reporting violent behaviour by guards and compiling past cases of abuse. 
AI has a much greater analytical range than humans, allowing it to study the depth of various elements in cohesion. For example, AI applications may consider a variety of factors such as an inmate's age, family history, native location, and the nature of the crime when allocating cells.

Artificial intelligence (AI) can help deter criminal activities and smuggling on prison grounds by detecting unusual behaviour and movement.

Prisoners from the same geographic pool are rarely held together in prisons because they are likely to have similar lifestyle patterns, making it impossible for them to avoid troublesome traits and stay on the recovery track.

If prisons must be correctional facilities, AI tools can be useful in preventing prisoners from straying, such as by aiding in the treatment of opioid addictions through surveillance. 
 
The Challenges
AI's limitations, such as inherent bias, opaque algorithms, and a lack of complete datasets, especially for the native community, can make the technology ineffective or even harmful. 
It is important to uphold individuals' human rights by using predictive policing techniques and technology in a democratic society regulated by the rule of law. As a result, by using analytics software, problems that are inherent to law enforcement agencies' analytical approaches could be avoided.

The threat of a loss of privacy and other human rights and democratic values such as the presumption of innocence and the prohibition of fines without a statute is one of the most critical areas under consideration.

In India, crime rates have risen dramatically over the last decade. According to the National Crime Records Bureau, the most significant increase has been in cognizable crimes, which has increased by about 63 percent. Departmentalized investigations are unable to provide law enforcement officials with a comprehensive image. A holistic review of various aspects of the knowledge is needed to counter and efficiently manage sensitive law and order situations. 
 
Artificial intelligence (AI) is sweeping the globe
Police forces all over the world have begun to focus on Artificial Intelligence technology in order to carry out policing functions more reliably and effectively. Artificial intelligence (AI) is constantly being used to detect suspects and search video recordings for anomalies in crowd control and surveillance. Simultaneously, the mere existence of such technology will serve as a deterrent to crime. It's as if the cops are still on the lookout for you. Artificial intelligence (AI) presents a wide variety of possibilities for transforming the way law enforcement is conducted. AI aids in the monitoring of large crowds and city-wide surveillance.

The New York City Police Department in the United States of America employs crime-prevention software such as CompStat.

In the United Kingdom, the technology that can be used for facial recognition, such as the CCTV surveillance system, has been introduced in Pembrokeshire.

Other Asian nations, such as Hong Kong and China, have begun to use AI to fix scale—linked sensors and monitoring wristbands to aid in the development of smart systems that could make prison breaks obsolete.

In Hong Kong, the government is experimenting with wearables to monitor people's whereabouts and behaviours at all times, including their heart rates. Some prisons, such as China's Yancheng prison, use networked video surveillance devices to keep tabs on high-profile inmates. 
 
Is it worth investing in Artificial Intelligence?
There are also some concerns that the industry and technologists must answer. Can the AI meet the Police's stringent criteria for speed, accuracy, and low error rates? Can we give the police one-of-a-kind proposals to help them save money on technology that shifts too quickly? Will we be able to have safe and dependable solutions that meet the needs of our future smart cities?

AI can not only assist in making footfall forecasts, but also in determining crowd density in real-time. AI simulations can identify 'choke' points and, based on moment trends, predict where crowding is most likely to happen, and, in some cases, can even help prevent a stampede from occurring. Slip and fall injuries can also be detected reasonably easily, allowing for the avoidance of more serious events.

Furthermore, the possibility of terrorist activity at a large event cannot be ruled out. Although police can make every effort to locate unusual or unattended items, AI has the hawk-eyes to spot such objects quickly and raise an alarm. It can also be used at railway and metro stations, where AI can mark an item with an owner beside it but no one to attend after a certain period. There is also technology that can detect fraud in a crowd. AI can quickly detect a firearm or gun being flashed outside a discotheque and automatically notify the authorities. 
 
The Concept of Predictive Policing 
Fortunately, the most recent tool for police forces in the fight against crime is data, not a powerful gun. Predictive policing is already in practice, and what matters now is stopping crimes before they happen. Data on the time, place, and nature of previous crimes is fed into statistical equations to give police strategists insight on where and when police patrols can patrol or maintain security to have the best chance of deterring or preventing potential crimes.

Many states in the United States, as well as countries such as the United Kingdom and the Netherlands, have successfully used data derived from population mapping and crime statistics, as well as current data, to make decisions, resulting in lower violent crime rates.

AI Predictive policing is the ability to predict where crimes will occur, who will commit them, what types of crimes will be committed, and who would be the victims. Predictive policing is a contentious topic, and it is still a long way from becoming widespread. Companies and police departments are only now beginning to put predictive policing systems to the test. These systems have the potential to make significant progress in terms of predicting and, ideally, preventing crimes.

The primary objective of law enforcement agencies is to provide society with a safe and secure atmosphere by preventing and solving crime. The changing social landscape necessitates that the police become better prepared to deal with crime.

We've witnessed police departments suffer from a force crunch in the past, like licenses, permits, and lengthy training periods stymie the process of new people joining the force, compromising its effectiveness, but with the right perspective, the police will be able to transform their cameras, Internet of Things (IoT) devices, and drones into a force that is on call 24 hours a day, 7 days a week, and information that never fades, as well as actions that develop with every picture it takes. 
 
Benefits of Predictive Policing

• AI can provide a more accurate view of who is a possibility of committing a crime and who is likely to re-offend after being released from jail, based on evidence and the history of the offender
• Predictive policing shifts the response process from responding to crime to forecasting the probability of crime and allocating resources to prevent it.
• It anticipates crime incidents and generates actionable information using predictive models and computational resources.
• Predictive analytics can be used to analyse variables like locations, individuals, classes, or events. It may also aid in the analysis of demographic patterns, parole populations, and economic factors that could have an effect on crime rates in specific areas
• It promises easy-to-access data collection, which will enable law enforcement officials to target policing by identifying individuals and locations. 
• It allows law enforcement officers to respond ahead of time to prevent crimes by concentrating on crime-prone areas and people that are at risk of offending or being attacked.
• Also, this is a new way to use AI to extract valuable information from long CCTV footage through fast real-time notifications, which cuts down on the time it takes to produce actionable data. 

The Challenges:

• Plan 
  There are insufficient precautions in place to avoid misuse. On a daily basis, predictive policing entails a preventive principle to the threat of crime. This preventative action also raises questions about inconveniencing innocent people and infringing on their rights. The Code of Criminal Procedure now allows for arrests based on suspicion. As a result, any misuse of the predictive policing system could result in arbitrary arrests and detention without justification.
   
• Confidentiality
  Although using data to identify hotspots or heat maps may not be a privacy problem, using data to identify likely individual criminals is. Any analysis of personal data can draw the attention of the general public and organizations. People are worried about the use of their data, and many of them may be reluctant to share details about their behaviour.
   
• Exaggeration 
  Data-driven decision-making processes tend to intensify existing information inequities. Any action to correct the information incorporates into the predictive policing data that guides decisions. Discrimination, which is a systemic prejudice, is a limitation on predictive technology.
   
• Algorithm in authenticity 
  In certain cases, courts are unable to understand predictive policing algorithms. As a result, existing statutory provisions aimed at preventing discrimination are ineffective.
   
• Ideology of data
  Predictive policing relies too much on data and lacks a slew of other variables. For example, areas highlighted by heat maps or hotspot analysis are considered for police patrolling while the rest of the city is ignored.
   
• Predictive models' incapacity
  What algorithm is written governs predictive models. It effectively draws attention to the fact that the data used, assumptions made, and the types of contextual questions posed by the algorithm are all completely unknown.
   
• Safety is paramount
  Based on an organization, the security of the data that will be used for conducting analysis and storing the reports after analytics is a major concern. It is important to recognize the need for a decent infrastructure facility for data safety and protection.
   
• Capture and storage of data
  Data can be accessed at a high rate from a variety of sources, and storing the captured data is a task in and of itself. After following the proper data protection and confidentiality protocols, data sources may include social media networks, mobile phones, weather forecast reports, websites, and other government entities such as the Unique Identification Authority of India (UIDAI), Crime and Criminal Tracking Network, and Systems (CCTNS), and National Crime Records Bureau (NCRB).
   
• An excessive dependency on technology 
  It's a common misconception that modern technologies can fix old issues. Technology, on the other hand, is merely a means to an end. Predictive policing systems can evaluate data, but it is up to the people who use them to interpret the results in a reasonable and just manner.
   
• Cybercrime is a growing issue
  Another big issue to be resolved is criminal development. Criminals value the data used by law enforcement agencies for predictive policing in order to deter and disrupt illegal activity because it allows them to commit more advanced cyber-enabled crimes. As a result, it's important to safeguard such information from cyber-attacks. 

 
The Facial Recognition Technique
Another field where law enforcement authorities are constantly using technology in their day-to-day operations is facial recognition technology. Face recognition is an essential part of policing. It is now relatively simple to distinguish offenders from large gatherings, whether in the form of a snapshot or a video clip. The officials only need a photograph of the person, which doesn't have to be recent or of high quality. Facial recognition is now being used in day-to-day policing in China. A body camera on a police officer's arm, for example, will tell him if the individual he's speaking with is on the police blacklist.

Face recognition undoubtedly has privacy implications, and police should exercise extreme caution when using it. While the Indian government plans to pass a privacy law soon, technology and regulation may be tailored to reap the benefits while staying within the bounds of the law and keeping the public's best interests in mind. 
 
Under Surveillance
It is particularly important in a country like India, where the population density is far higher than the global average, not only in large cities but also in smaller towns. Crowd events are one of the most difficult obstacles that any city police force faces, and we can no longer rely on conventional crowd management techniques.

Video analysis is difficult due to the nuances of obscuration and inconsistencies in crowded situations. Although cameras at a crowd gathering and drones flying overhead can help, the use of AI-based technology that has learned crowd behaviour completely changes the situation, allowing police to anticipate behaviour and make more rational choices. 
 
Rise of the planet of the AI bots
In reality, some countries are experimenting with robots that can replace police officers. Dubai is testing street robots that can relay data to headquarters, where it will be checked by humans. They also have touchscreens that can be used to track offenses and can communicate in six different languages.

More complex problems can also be completed by robots on behalf of cops. They will access dangerous areas and recognize people and items that could pose a threat, which is a better alternative to risking the lives of police officers. There are now robots with the potential to detonate explosives, enhancing public safety without placing officers in harm's way. 
 
Positive Development: A Threat to Human Rights
The use of artificial intelligence (AI) in policing is on the rise. This productivity may also come at the expense of human rights, which is a price that we as a society might not be willing to pay.

Because of the inherent ability of police to seize, prosecute, and even use lethal force, it is important to recognize that adopting AI in policing would have different and much more serious implications for human rights. Governments all over the world are developing artificial intelligence (AI) strategies for governance and administration without taking into account the effects on human rights.

Without a question, technology must be used to improve productivity in all sectors, including law enforcement, but there is a pressing need to consider the implications of AI use and to develop strategies to mitigate the damage it causes. Before the introduction of these invasive AI instruments, a risk evaluation must be completed.

To the technology's detractors, it appears that law enforcement agencies are developed enough to comprehend the technology. They know how to get the most out of technology while still being cautious and not becoming too dependent on it. Human factors, as the ultimate decision-maker, should not be overlooked. Technology is here to stay, and the only question now is how smart police can use it to keep our smart cities secure.

Various Artificial Intelligence technologies are being used by police forces all over the world to aid in human decision-making. With the increased implementation of AI systems in various industries, several human rights issues have been raised. On the one hand, countries are filing complaints against law enforcement agencies for violating citizens' privacy, while on the other hand, some countries are indiscriminately employing AI at the expense of their citizens' human rights.

As a result, it is critical that the consequences of AI technologies be closely monitored and thoroughly analysed before they are implemented. 
 
How the Indian States are using Machine Learning in Policing
The National Crime Records Bureau (NCRB) has been working on a program to use crime data for analytics purposes to facilitate predictive policing in collaboration with Hyderabad-based Advanced Data Research Institute (ADRIN). After taking into account different variables and conducting research, the project uses deep learning algorithms to forecast geographic areas where crime is most likely to occur.

Police will schedule patrolling, resource allocation, and surveillance based on the findings, ensuring that no crimes are committed. 'Hotspot analyses', where law enforcement authorities can analyse and forecast geographical areas of heightened crime based on crime data, is one of the project's standout features. All trends can be processed using hotspot analysis, regardless of the time of occurrence, individual locations, or even stores, hotels, bars, or other establishments.

The Hyderabad Police Department has also signed a memorandum of understanding (MOU) with 'Synchrony Financial' for the insertion of a community closed-circuit television (CCTV) surveillance system in the areas. The installation of IP-based outdoor security surveillance cameras, automatic number plate recognition (ANPR), video analytics, mobile surveillance system, command, control centre, and data centre, among other things, is part of the CCTV-based video surveillance system. The proposed structure would include a multi-agency operation centre, as well as a location for each department's technology teams to review CCTV footage and receive input to motivate their respective field operation teams.

In 2017, the Delhi Police collaborated with INNEFU Labs' AI Vision facial recognition program, which provides gait and body analysis. INNEFU, a domestic artificial intelligence start-up, is capitalizing on India's burgeoning market for facial biometrics by conducting tests on Indian faces at affordable rates.

Through a new program called CMAPS, it has been improving the ability to recognize crime hotspots and reduce the likelihood of any repeat crimes in the area (Crime Mapping Analytics and Predictive System). The web-based program has real-time access to information from the Delhi Police's Dial 100 helpline, and it can temporally locate the calls and visualize them as cluster maps using ISRO satellite imagery to locate crime hotspots.

Similarly, police in Odisha are preparing to use artificial intelligence and mobile computing to boost crime data analysis. Correctional officers would be able to use AI to detect procedural errors. Odisha Police issued a request for proposals (RFP) in December 2019 to solicit bids for qualified AI applications.

Jharkhand Police:
The Open Group on E-governance (OGE), which was established as a result of a partnership between the Jharkhand Police and the National Informatics Centre, is a multi-disciplinary group in charge of various information technology projects in the state. There have been proactive attempts to improve expertise in predictive policing. For example, OGE developed data mining tools that would be able to search digitized online documents. These are expected to serve as building blocks for the state police's proposed predictive policing initiative.

Maharashtra Police:
The Maharashtra government is working to update its digital technology-based policies to incorporate "predictive policing policy" as part of its cybersecurity modernization program. The initiative is expected to aid law enforcement authorities in predicting, preventing, and detecting cybercrime.

In addition to legislative changes, the state government will establish a new department named MH-CERT, which will be similar to the Centre's Computer Emergency Response Team (CERT).

The department will reduce the state's reliance on the federal government to deal with situations where social media is used to spread misinformation that lead to law and order issues. Police may use predictive policing to monitor social media data in real-time to identify individuals who are attempting to incite violence.

Kolkata Police:
Based on data obtained from Google maps, Kolkata Police have proposed a framework that uses analytics to efficiently control city traffic and maximize the number of vehicles passing via intersections. The proposed system's goal is to predict the exact signal period using real-time traffic data from Google Maps and then calculate traffic lines at intersections around the region. All of the signals in the city are linked via Wi-Fi, allowing for central signal management. 
 
Conclusion 
Artificial intelligence in law enforcement has emerged as a critical component of police work around the world. Due to the overloading of jobs, India's police force often faces health and social problems, necessitating the need for better resource distribution. Police officers typically work seven days a week and are often required to work very long hours. As a result, any technology or police infrastructure that allows for efficient resource distribution is highly desirable.

Areas like crime prevention and prediction are undergoing dramatic changes as AI-based policing technology is becoming increasingly relevant to law enforcement. Other police strategies have undergone major changes in the name of public safety, and predictive policing is only one of the outcomes of this transition. If crimes can be prevented before they are committed, it has significant social and economic benefits not only for those who are at risk of becoming victims but also for the perpetrators, who can avoid making life-altering errors.

Before implementing AI, it is critical to establish a secure atmosphere for its implementation and to recognize the risks. It is necessary to determine the effect of this technology and implement policies to prevent the damage that it will cause to the human rights regime. The advantages of technology can and will be maximized only if attempts are taken to mitigate the harm that this disruptive technology can cause. 
 
References:

• https://cpr.unu.edu/publications/articles/ai-global-governance-turning-the-tide-on-crime-with-predictive-policing.html
• https://bprd.nic.in/WriteReadData/CMS/Policing%20in%20Smart%20Cities.pdf
• https://government.economictimes.indiatimes.com/blog/to-get-best-out-of-technology-indian-police-must-ditch-the-silos/3972
• Policing in the Era of AI and Smart Societies by Jahan Khani, H., Akhgar
• Special Issue on Artificial Intelligence for Cyber Defence and Smart Policing, http://www.wikicfp.com/cfp/servlet/event.showcfp?eventid=101581©ownerid=101185

Written By: Aditi Chauhan, Final-year BA.LLB student, FIMT School of Law, Guru Gobind Singh Indraprastha University