“Policing by consent”: Modern British Policing and Live Facial Recognition

How the adoption of surveillance technologies is challenging British policing’s long-standing model of consent

The long-standing model of British policing is “policing by consent”, meaning that the power of the police is derived from the respect and approval of the public more than from the authority of the state. It denotes that the legitimacy of policing in the eyes of the public is based upon ‘a general consensus of support that follows from transparency about their powers, demonstrating integrity in exercising those powers and their accountability for doing so.’

Unlike consent in commercial settings, consent to the authority of law enforcement must exist collectively, as opposed to individually. The government states this clearly: “[Policing by consent] does not mean the consent of an individual. No individual can choose to withdraw his or her consent from the police, or from a law.” This raises a host of questions with regard to the government’s definition of the word ‘consent’- if consent cannot be withdrawn, can it be considered consent at all? Can British policing today – particularly when we consider the use of increasingly controversial technologies with little to no public input or ability to opt-out – be considered truly consensual?

This centuries-old concept is being challenged, with innovations in technology forcing policing to adapt and respond to changing demands. New methods of crime prevention, investigation and evidence gathering have been integrated into police forces across the UK, and the need for this has been highlighted by Sir Tom Winsor, Chief Inspector of Constabulary, the independent regulator of policing in England and Wales:

“It is essential that the police are given the means to [invest in new technology]. For example, body-worn video, facial recognition and artificial intelligence ... If they don’t, they are left playing catch-up as offenders abuse modern technology to cause harm.”

One of the most controversial developments has been the adoption of Live Facial Recognition Technology (henceforth referred to as LFR) as a surveillance and identification tool. This controversy stems from well-documented concerns as to the neutrality of this technology, as LFR technologies are notorious for behaving differently depending upon an individual’s sex or ethnic background, thereby giving rise to discrimination. These disparities are built into the tools themselves, arising from the manner in which the neural networks that generate the algorithms are designed, and from the datasets on which they are trained.

A 2018 study by MIT and Stanford found an error rate in LFR technologies of 0.8 per cent for light-skinned men and 37.4 per cent for dark-skinned women. Likewise, a December 2019 report from the US National Institute of Standards and Technology surveyed 189 commercial algorithms from 99 different developers - a majority of the industry, and found similar discrepancies.

And the police are not unaware of this. In fact, in a report on the London Metropolitan Police Service’s Trial of LFR, this was specifically raised as a concern regarding the MET’s adoption of LFR, with the report citing that ‘[LFR] is analogous to traditional concerns associated with potentially discriminatory policing practices.’ The implication here is that the adoption of LFR will entrench discrimination in British policing, as opposed to working against it or even existing independently of it. These dangers are a real public hazard, and it is not inconceivable that the public would be opposed to their police using such tools, were they given a choice.

CCTV and “surveillance by consent”

Any discussion regarding the broader ideas of surveillance and consent in British policing would be remiss in not considering the history of Closed Circuit Television (CCTV) cameras in the UK.

Analogous to policing by consent, “surveillance by consent” as a concept first emerged in the Surveillance Camera Code of Practice issued under the Protection of Freedoms Act 2012. This code was introduced to “further regulate” the use of CCTV cameras in public spaces, as well as to clearly outline the notion of consent in the context of surveillance technologies.

“Overt surveillance in a public place can be said to have been consented to if it is deployed in pursuit of a legitimate aim and meets a pressing need, and when the deployment is proportionate to the stated purpose without unnecessary interference with privacy and other human rights protected under the Human Rights Act 1998.”

However, existing legislation regarding “surveillance by consent” is widely regarded as incomplete at best, with Privacy International stating that it ‘advances the deterioration of protections that were already in place.’ This raises concerns regarding attempts by police forces in England and Wales to deploy LFR technologies, particularly that we will simply be replicating the key issues with the exponential deployment of state-funded CCTV cameras in in the 1990s.

To demonstrate: the earliest form of code of practice for CCTV was a code published by the Local Government Information Unit, which stated: “No sound should be recorded in public spaces”. Now, the new Surveillance Cameras Code of Practice states: “Any proposed deployment that includes audio recording in a public place is likely to require a strong justification of necessity to establish its proportionality”. We have moved from a clear prohibition to a vague statement that justifies its permissibility, with little to no public backlash, raising concerns as to the future of state-sanctioned surveillance technology.

LFR and Earned Consent

When considering consent, it is interesting to note that there is currently no law explicitly authorising the use of LFR in the UK.

In the absence of an explicit legal framework, police rely on previous legislation, including common law, the Human Rights Act 1998, the Protection of Freedoms Act 2012, the Data Protection Act 2018, and the Regulation of Investigatory Powers Act 2000. Therefore, it is justified to say that in the case of LFR, consent is assumed, not sought, similar to the implicit consent given to CCTV.

In fact, the Scottish Parliament has effectively called for a pause in police use of LFR, demanding that Scottish police demonstrate a legal basis for the use of the technology, including its compliance with human rights and data protection obligations.

As well as legal challenges, there has also been a significant increase in public scrutiny since the deployment of CCTV in the 1990s, with concerns -in particular regarding the potential of racial discrimination- gaining traction through social media. In fact, research by the Ada Lovelace Institute shows that 46% of respondents agreed that the public should be given the opportunity to opt out of being subjected to facial recognition technology. Therefore, consent should no longer assumed, but earned.

As Foucault observed, governmental technologies are never simply oppressive. They are productive, and liberal governmentality is an exercise in balancing productive power against its counter-productive risks. And under the notion of “policing by consent”, the public must be made aware of these risks.

Real choice demands a wider assessment of surveillance technologies, both for existing and new technology. Real consent would require the public to be informed about the dangers of technologies with built-in biases. It would require a meaningful public debate about whether the benefits of LFR systems are worth the dangers. And so it cannot be justifiably said that British policing today is truly consensual.