AI Ethics in Law Enforcement: Navigating Innovation and Oversight

The shiny facade of the recent ‘global AI safety summit’ doesn’t offset the government’s serious lack of ambition on ‘safety and ethics’ in the UK. This broad ambition to lead the diplomatic effort on AI, is honourable in its own right. Dystopian depictions of our future, envision AI attacking our vital infrastructure, turning on its creators and exterminating humanity. Realistic or not, the fear of Armageddon has inspired international cooperation.

This government has so far displayed a worrying lack of urgency on AI safety particularly in the area of law enforcement. Framed as unnecessary bureaucracy or ‘slowing progress’ AI ethics in policing and justice is seen as an inconvenience. While Prime Minister Rishi Sunak pretends to lead the international agenda, he neglects to get his own house in order; potentially allowing the immediate rollout of harmful technology in our communities.

Innovation in AI could prove crucial in rebuilding the UK’s enforcement capability in a decade in which serious crime going unsolved is the new normal. The wheels of justice have all but ground to a halt over a decade of austerity. Cyber fraud sees thousands of British citizens scammed out of billions of pounds annually. Then there’s the appalling injustice of low sexual offence prosecution rates – only 3.6% ended in a charge in 2022-23. As ex-Victims Commissioner Dame Vera Baird says “The whole process has been more traumatic than the actual rape. I have zero belief in the justice legal system.” We need modern answers. Exploring how AI can apprehend dangerous people, prevent serious crime and promote effective prevention, should be a priority. Indeed, ChatGPT tells me crime-busting AI tools include Predictive Policing, Facial Recognition, Natural Language Processing, Crime Analysis and Pattern Detection, Prisoner Classification, Risk Assessment, Evidence Analysis, Chatbots and Virtual Assistants and Predictive Analytics for Officer Deployment. This is a fair reflection of what law enforcement bodies are currently working on.

Just Get It Out There

Predictive policing represents one area in which AI could enable a more effective allocation of policing resources and begin to turn the tide on the policing failings under the Tories. Controversially, predictive analytics seek to apply patterns using data from past events, to predict future crimes. We’re nowhere near producing accurate or reliable predictive policing, especially when it comes to individual offending of victimisation. The data quality, information sharing and clarity of objectives, just aren’t there yet. But facial recognition technology is here already. This could help tackle human traffickers and identify dangerous individuals. But I don’t want it to discriminate against women or people of colour as some fear. Nor do I want to see it aid the punishment of young people, who are better supported through a reversal of a generation of youth service cuts. The truth is, the government takes a “just get it out there” approach. Recently, Police Minister Chris Philip urged police forces to double the use of facial recognition software despite concerns over its use by a cross-party group of MPs and Peers. The lack of parliamentary scrutiny and oversight over a measure that has significant implications for citizens’ human rights, data privacy and police enforcement capabilities is jarring. Just like its strategies for green technologies, rather than be the world leader in ‘AI ethics and safety’, Sunak would rather take shortcuts than lead.

We need the government to put ethics at the centre of AI innovations in policing. Proposals that AI tech companies should robustly demonstrate their ethics safeguards, are met with Ministerial shoulder shrugs. Currently, as with facial recognition technology police forces can experiment without proper oversight and the government is dragging its heels over recommendations from the House of Lords to establish a truly independent ethics body for law enforcement. There are some brilliant senior officers who really understand that we need to embed a culture of ethics urgently, not just for today’s tech, but for the next 100 years. There are others who are prepared to take the risk, as though the damage in public trust to policing in recent years hasn’t been enough. If we end up with technology that automates racist criminal justice decisions, the government cannot say they haven’t been warned. We need a truly independent national ethics body for policing capable of communicating honestly to the public; without political interference and lobby interests. And yet rather than show leadership on the actions needed in the UK, we’ve a Prime Minister telling the rest of the world to do it instead.

A Fresh Start

Labour recognises that AI paves the way not only to a safer society but a healthier, greener and more productive world. It can be a source of highly skilled jobs and responsible growth. Vitally, Labour also has the moral backbone to make ethics a business advantage and not an afterthought – making it central to the country’s technological ambitions.

For more on A.I see Missed Opportunities in the A.I Summit.

Author

  • Tom McNeil

    Labour & Coop Assistant Police & Crime Commissioner for the West Midlands Policy Fellow University of Cambridge Centre for Science & Policy Senior Visiting Practitioner University of York Administrative Fairness Lab

    View all posts