Law Enforcement Agencies
As the world gathers this weekend in Singapore for the Annual Interpol World Congress to discuss law enforcement cooperation, we are all reminded of the impact of AI/IoT/Cloud/Big Data technology in advancing law enforcement. The policing landscape is changing very fast: the summary below looks at covering some of the emerging trends in law enforcement.
AI as Sherlock
In May 2017, the New Scientist published an article on VALCRI, an EU initiative that puts AI at the heart of crime analyst.
VALCRI will be able to take historical data from disparate sources, analyze it and provide suggestions into why a crime was committed, the likely guilty parties, the timing and method used for the crime. With the help of user friendly visuals, VALCRI doesn’t replace police detectives completely but helps in narrowing down possibilities or bringing unlikely information together at times. According to Middlesex University (who are partners on VALCRI) an experienced analysts used to require 73 individual searches but now can do the same work with one click. Some police stations in the UK and Belgium are testing the software and whilst some have raised concerns over the impact a predictive software would have on discriminating against individuals others are hailing its prospects.
Extremism detection
In the US, NSA and NGA continue to improve their AI capabilities to go through millions of messages and images to detect potential threats. The Stanford University study, Artificial Intelligence and Life in 2030 report revealed that some law enforcement agencies are using AI to monitor and analyse communication on social networks to detect hate speech and attempts of radicalisation. iAWACS is one example tool that has been used in that space.
City Surveillance:
Fog computing enables smart camera to improve city surveillance. Today most cameras are used to understand a crime after it happened rather than prevent it. With the ability to extract the right information and make decisions we could see an increase in the number of applications. One that caught my eye was drone surveillance used in India to control crowd in 2015. The drone was equipped with pepper spray, shooting paintballs, speakers, cameras and microphones and was seen as being a more effective method of enforcing order without putting police at risk. Whilst these drone where being remotely controlled there is definitely work underway to improve their ability to recognize threats and react to them independently. Last month India signed a contract with the US to purchase $2 bn worth of drone technology to improve maritime surveillance over the Indian Ocean.
Cybersecurity
The last few years have highlighted how large companies, governments, private individuals are vulnerable to cyber attacks (e.g., National Health Service in the UK, American election alleged hack). Securing federal networks, protecting critical infrastructure, defining appropriate incident response are all part of the work that is underway. The staffing profile of these agencies is slowly transforming to balance the need of traditional law enforcement workers with savvy security engineers.
Smart Guns
In the wake of various police shootings in the United States, officials are looking at ways to mitigate the risk of police officers shooting at low-risk suspects. One of the solutions proposed is equipping guns with “IoT components such as GSM and Bluetooth chips as well as motion sensors that can provide information useful to ballistics experts and forensics investigators. The data is uploaded to IoT servers where a quick algorithmic assessment of the situation can be conducted and a signal can be transmitted to the officers on the scene by means of the machine-to-machine (M2M) protocol.”
Whilst all the above enhancement can allude to a Minority Report like reality, where crime can be averted, we are still far from a perfect word. For one a lot of work needs to be done to strengthen AI capabilities and more testing needs to be carried to improve performance of IoT devices. Also many of these technologies can be used by perpetrators to weaken to system. The fight between the light and dark side of the force has just reached a new level.
References
- AI detective analyses police data to learn how to crack cases, New Scientist, May 2017
- One Hundred Year Study on Artificial Intelligence, Stanford University , September 2016
- Flavio Bonomi, Rodolfo Milito, Jiang Zhu, Sateesh Addepalli , Fog Computing and Its Role in the Internet of Things, Cisco Systems Inc. 2012
- The Internet of Things is Changing the Future of Law Enforcement, www.iotevolutionworld.com, September 2016
- Trump Approves $2 bn sale of drones to India Ahead of meeting, The Washington Times, June 2017
Users who have LIKED this post:
5 comments on “Law Enforcement Agencies”
Comments are closed.
Great perspective on one of the most pressing issues recently. I have many friends who are in various federal and local branches of law enforcement, and the topic of Bluetooth technology as it relates to ballistics is of extreme interest to them. A firearm with a digital footprint could be helpful in forensic analysis, but I wonder if 2nd Amendment advocates would push back citing privacy laws. Great post!
Interesting article to read Lana.
Nevertheless, it is necessary to say when comes to cybersecurity attacks as in the case of NHS in the UK, there is very often a lack of incentives on the user side to ensure the highest possible security. After recent cybersecurity threats, it was investigated that 90% of NHS in the UK were using Windows XP (16-year-old operating system) which is particularly vulnerable and causes a higher risk.
Hey Lana,
I’d first like to say, a great read and the concluding paragraph was creative and entertaining. A few comments on a few things.
I found the VALCRI an interesting initiative as it ties well to some of the trends highlighted by Cruz. Law enforcement data across the US for different counties, jurisdictions and states are all siloed off and I can see huge societal benefits in aggregating and using it for these types of applications. Up until now, the majority of these types of initiatives are spearheaded and implemented by governments partnering with private companies. Cruz mentioned in his presentation that California is beginning to make large amounts of data public. What would be interesting to see is if governments at the federal level would open source large amounts of past criminal data to private companies to use it. An example of this would be potentially releasing data on homicides in the entire country and allowing private companies to leverage analytics to identify potential suspects for a bounty. I’m not sure what the company’s name is but I did read about a company doing something similar to what I just outlined, if I have come across it again I’ll let you know.
Michaela- I think Cybersecurity will need to transform in the coming years to address that issue. Today, a security breach needs to happen before you realise you have to update your software and whilst a lot of white and black hat testing happens it is not enough to fully protect consumers. I hope security engineers will develop warning signals so that we can get alerted earlier on before an attack happens to prevent it.
Johnny-whilst I think it is possible law enforcement agencies would share data I am not sure if they will be able to share information to identify victims- as this maybe against data privacy laws unless victims explicitly consent to it.
Rob- thanks for your comment. I am with you on the privacy laws piece.