21 Mar 2022

Policing in a digital age - how Artificial Intelligence, Intelligent Automation and Cyber can help keep citizens safe

Neil Richardson, Solution Architect, Policing & Justice at Kainos as part of techUK's Emerging Tech in Policing week

The nature of policing has changed in recent years as digital technologies have been exploited by criminals. 90% of crime now has some digital element, and advances in technology enable novel crimes or the means to incite civil unrest, adding to the challenges that police and public protection agencies already face.

At Kainos, we believe that police services around the world must adapt the way they traditionally police, embracing innovative technologies so that policing stays ahead of the curve and is future-proofed for years to come.

From my experience of working in the sector, I believe that the three key technologies important to policing in the next 10 years will be artificial intelligence (AI), intelligent automation (IA) and cyber security. Here I’ll explore these and explain how we at Kainos help police services to adopt these latest technologies to build coherent and effective policing solutions.

Data, Artificial Intelligence and Machine Learning (ML)

Volumes of data collected by the police are already large, and the explosion of new sources of data such as video, smart cities, internet of things, and social media means police services will need to rise to the challenge of understanding, processing, and making use of exponentially increasing volumes of data.

This volume, while technically challenging, presents opportunities around AI and ML to reduce the burden on users, providing opportunities to operate at scale, in a reliable manner, and remove instances of human error.

Our work within policing and public protection, demonstrates how AI can be used in the detection of crime, building intelligence profiles and networks of information, and improving efficiencies of operational functions.

Of course, one of the challenges with AI services is gaining public trust; trust that it will function correctly, be reliable in real-world conditions, be legal and ethical, and suitable for its intended purpose - which is why we are hiring a Data Ethicist.

As well as establishing this vital role, our finely tuned risk management approach for AI delivery recognises these challenges and uncertainties in AI regarding data availability, ethics, live deployment, and human-in-the-loop decision making.

Our successful AI delivery partnerships have helped to refine this approach. To accelerate and de-risk delivery, we:

  • use and license our substantial, pre-existing, and proven AI intellectual property for policing use, including our MLOps Framework, AI Safety tooling and AI Education materials.
  • use frameworks such as V3 (Verification, Validation and Vulnerabilities) to keep track of the overarching objectives that need to be considered when designing, building, and productionising AI.

Intelligent Automation to free up specialists

When robotic, intelligent, and autonomous systems are integrated, the result is IA, creating scope for tasks and processes to be automated and the burden on staff to be reduced.

There are several potential use cases of using Robotic Process Automation (RPA) in policing which include:

  • Performing repetitive tasks like data entry, freeing up staff to perform higher-value tasks, improve quality and remove bottlenecks (as robots can be easily scaled up to handle peaks in demand).
  • Providing automated integrations between diverse legacy systems offering immediate value without the risk and expense of large systems change.
  • Intelligent Document Processing – Machine Learning (ML) to digitise, classify and extract meaning/content from documents. 
  • Chatbots to triage and respond to simple enquiries from the public and handing-off appropriately if unable to handle itself.
  • Speech-to-Text to transcribe voice records from telephone/mobile/radios, allowing staff to just review output rather than conduct full transcriptions. 

Of course, Intelligent Automation is not without its own challenges - secure implementation, due to handling sensitive information, need to be in place, and you need to gain buy-in at all levels, with appropriate controls to help realise the maximum benefits of this technology.

Cyber as a key enabler

In a policing context, cyber security becomes a key enabler to the delivery of innovative technical solutions, such as AI, Machine Learning, and Intelligent Automation, to manage and analyse large volumes of potential sensitive data which would traditionally have resided within classified or air-gapped networks.

Public cloud providers are increasingly offering best-of-breed out-of-the-box accelerators supporting AI and IA, with accompanying cyber security technologies and practices - for example, moving to zero trust networks ensuring that data is safe from attack and loss within cloud environments, and able to be transferred securely between networks.

Our dedicated Cyber Security practice advises our delivery teams on best practice and can also enable this change. We accelerate the adoption of cloud by building security into the heart of DevOps processes. We advise on and implement development and operations processes, including:

  • Security by design, in high-level/low-level design artefacts 
  • Applying application security controls consistently via OWASP ASVS (Application Security Verification Standard) framework, e.g., protecting user inputs from malicious manipulation
  • Applying infrastructure security controls consistently via CIS (Cyber and Information Systems) benchmarks and vendor specific security recommendations.
  • Shifting left, catching vulnerabilities early and implementing automated security testing in delivery (CI (Continuous Improvement)) pipeline (e.g., SAST (Static Application Security Testing)/IAST (Interactive Application Security Testing) tools, secrets checking, dependency scanners)

So, what will policing in a digital age look like?

We believe that Artificial Intelligence and Intelligent Automation will provide policing with a solid foundation into the next decade. It will allow faster and better decision making, improved deployment of skilled frontline staff and help police to work smarter. Accelerated through cloud adoption and underpinned by cyber security principles, we’re confident that the challenges may be easier to overcome than they might appear.

With our proven background in applying these technologies at scale, in addition to a long history of delivering successful digital transformation across central government, justice, and the police and public protection sector, Kainos has the people, tools, knowledge and passion to enable policing to achieve even more.

Find out more at www.kainos.com

Further reading:

Artificial Intelligence

Intelligent Automation

Related topics