A question of public trust

Scepticism about the use of artificial intelligence (AI) is a growing concern in London policing, particularly as concerns about the increase in AI facial recognition to catch criminals shape the debate over privacy. Most recently, the Mayor’s Office for Policing and Crime’s refusal of signing the £50 million contract for the Metropolitan Police Service (MPS) with US software provider Palantir sparked public criticism over the force’s use of general AI. However it’s a pivotal part of a criminal investigation.

Our mobile phones capture everything – where we go, who we interact with, and when it comes to investigations, that data is often the critical piece of the puzzle, helping to exonerate the innocent and place the guilty party at the scene of the crime.

Digital evidence is increasingly where police investigations begin, with 97 per cent of criminal investigations involving a mobile phone. The volume of data has exploded alongside the use of digital devices, yet the number of investigators to comb through that evidence remains small, or in some agencies, is decreasing. This has led law enforcement agencies to explore different forms of technology, such as AI, to manage and stay on top of this.

However, the relationship between the public and the police is fundamental and civil liberties must be at the forefront of decision-making for officers and staff.

How AI is changing modern policing

There’s a growing data problem in policing, which is placing pressure on law enforcement to modernise investigations for speed, efficiency and public safety. Public trust and legal protections have been a mounting concern for the public over their civil liberties.

An average personal device contains hundreds of thousands of messages ranging across sources from phones, cloud, social media platforms and encrypted applications, which will only increase as data increases into terabytes, causing more data to analyse.  As investigators search through the sheer volume of digital evidence, more equipped tools must be deployed to not only cross-correlate across multiple storage formats but also analyse and take action on any lead that comes from it.

AI capabilities have been deployed into forensic analysis to meet this challenge and support law enforcement workflows to speed up evidence analysis and improve efficiency. Using Generative AI in particular gives officers the ability to generate contextual summaries from datasets and act as a force multiplier in advancing and solving cases.

City of London Police and the MPS use AI to catch criminals in hundreds of thousands of cases each year and serve justice, with digital evidence sitting at the centre of major investigations.

It’s important to note that while AI can support police officers and investigators, it does not replace human judgment, and this must be present at all points of an investigation to ensure fair and accurate outcomes.

The policy gap

AI is being adopted in multiple industries, and policing and law enforcement is no different. The shift towards technology requires a shift in governance frameworks and how public bodies ensure they remain accountable and transparent to the public.

AI has been brought into digital investigations to automate the summarisation of evidence and sort and detect patterns, which frees investigators to focus on more interesting tasks, which can connect evidence across sources in minutes instead of days.

Law enforcement is continuing to progress with AI, but the main priority is ensuring this doesn’t infringe on human rights. Communication, therefore, becomes the focus between government, regulators and innovators to make sure this is at the forefront of decision making.

The growing debate around privacy

Like with any use of AI in public sector bodies, the public must be aware of how it is used, what this means for the individual’s data and what the appropriate safeguards are that are in place. Any use of AI and automation in policing should follow consultation with the public and be applied responsibly and where appropriate.

Police forces are committed to respecting privacy and ensuring the protection of personal information. Strict security and compliance measures are in place to abide by global privacy laws.

What needs to be focused on now is a more deliberate approach that defines and easily explains to the public what AI is used for, how decisions are reached and where the human sits in the process of this.

AI in policing is only one part of the broader picture, and the reality of what it should look like requires the technology to be used in a manner that is transparent and accountable. To get to this point, policy frameworks should be at the centre of AI innovation within the police and provide an honest outlook on where the boundaries of AI are.

John Lucey is VP EMEA North of Cellebrite.