SJUK hears exclusively from Tony Porter, Chief Privacy Officer of Corsight AI about integrating AI into modern day Policing.
Having spent much of the last decade involved in the public debate surrounding facial recognition, I have often felt that we have been asking the wrong question.
The question is rarely whether the technology works. Increasingly, it does. The more important question is whether the public can have confidence in how it is being used.
Artificial Intelligence is already transforming policing. From facial recognition and image analysis to the examination of vast quantities of digital evidence, AI offers capabilities that would have been difficult to imagine only a few years ago.
At a time when policing faces increasing demand and finite resources, it would be irresponsible not to explore technologies capable of improving operational effectiveness and protecting the public.
My experience, both as a police leader and later as the UK’s Surveillance Camera Commissioner, taught me that technological capability alone is never enough.
During my time as Commissioner, I was closely involved in many of the debates that shaped policing’s approach to facial recognition.
I appeared as amicus curiae in the Court of Appeal during the Bridges litigation and authored Facing the Camera, the first regulatory guidance for law enforcement on the lawful and ethical use of facial recognition technology.
Those experiences reinforced an important lesson not that policing should be wary of innovation but that innovation without governance invites challenge, whereas innovation supported by transparency, accountability and oversight earns legitimacy.
Chief Constables are not judged solely on operational outcomes.
They are expected to demonstrate that the technologies they deploy are lawful, necessary and proportionate.
Their decisions are scrutinised through Data Protection Impact Assessments, human rights considerations, ethics reviews, community engagement and your daily newspaper.
The scrutiny is unrelenting.
One lesson I learned during the facial recognition debate is that legitimacy cannot be retrofitted.
Once confidence is lost, rebuilding it is significantly harder than establishing it from the outset.
That is why governance matters.
Legal Eagles amongst you will have gathered that from the recent High Court Case Thompson and Carlo v Metropolitan Police.
Effective governance is not an administrative exercise or a compliance obligation, it is the foundation upon which public acceptance rests.
The reality is that policing is increasingly being asked to assess technologies whose complexity extends far beyond traditional procurement models.
Forces are expected to understand not only what an AI system does, but how it has been developed, tested, governed and monitored.
Frankly, that is asking a great deal of policing. It is too much to believe that this is possible
Chief Constables can and should be accountable for the lawful deployment of technology.
What they cannot reasonably be expected to become is expert auditors of every AI provider from whom they procure solutions.
Few forces possess the resources or specialist expertise required to interrogate every aspect of a supplier’s development processes, model governance, risk controls and lifecycle management arrangements.
The challenge for policing has never been choosing between innovation and public confidence.
The real challenge has always been finding a way to secure both simultaneously.
In my view, that is precisely where ISO/IEC 42001 becomes relevant.
Too often, standards are viewed through the narrow lens of compliance.
That risks overlooking their broader value.
The significance of ISO 42001 is not that it creates another layer of process.
It provides independent evidence that governance, accountability, oversight and risk management have been embedded within an organisation developing or deploying AI systems.
For policing, that changes the nature of the conversation.
Rather than attempting to independently assess every aspect of a supplier’s internal governance arrangements, forces can obtain greater assurance that recognised standards of accountability have already been subjected to external scrutiny.
Equally, responsible suppliers have an opportunity to demonstrate that governance is integral to their business rather than an afterthought designed to satisfy procurement requirements.
Importantly, it helps align public expectations with operational policing needs.
The public quite rightly expects transparency, accountability and safeguards whenever powerful technologies are deployed.
Policing quite rightly expects suppliers to demonstrate those same principles.
ISO 42001 creates a framework through which those expectations can begin to converge.
That is why I believe its significance extends far beyond procurement.
The NPCC AI Covenant and wider AI Strategy have already established the principles that should govern the future use of Artificial Intelligence in policing.
Demonstrating adherence to them is the next step and that is critical in being able to engage the public to demonstrate trust.
Trust, accountability, transparency and oversight are no longer aspirations; they are essential requirements for maintaining legitimacy.
The challenge now is ensuring those principles are reflected consistently throughout the ecosystem that supports policing, including the organisations that develop and supply these technologies.
If the NPCC strategy defines the destination, ISO 42001 may provide the most credible route currently available for getting there.
This should be a fundamental’ stop-go’ gate in any procurement of AI technology as all suppliers with the accreditation have had to demonstrate how they provide all the steps linked to the covenant.
This standard must now be mandated by the government for all AI solutions so that the Covenant can be adhered to.
From my perspective, having seen these issues from both operational and regulatory standpoints, that is the real significance of ISO 42001.
Not as a compliance exercise, nor as a badge of assurance, but as a practical framework capable of aligning public expectations, policing responsibilities and supplier accountability.
The future of AI in policing will not be secured by innovation alone, it will be secured by the confidence that innovation is being governed responsibly.
We congratulate NPCC for being bold enough to publish a covenant, to back this up now is the time to mandate ISO 42001 Standard to underpin and protect that covenant
