Simon Randall, CEO and Founder, Pimloc evaluates AI technology and surveillance in British policing.
The use of video surveillance technology has long placed British policing at the centre of a debate around transparency, trust, security and privacy. Now, the UK Government’s new plan to deploy facial recognition technology (FRT) with body-worn video cameras (BWCs) in law enforcement has further loaded pressure on authorities to deploy the next-generation of security technologies in a responsible way.
BWCs help substantiate or refute allegations of improper conduct. Trust and confidence in policing is also supported by allowing wider sharing of BWC footage with local communities, facilitating transparency and scrutiny of policing activities.
Early research shows that BWCs can positively impact relations between police and the communities they serve. One study by Cambridge University found that officer use of BWCs in the UK and US resulted in a 93% decrease in complaints filed against the police by members of the public.
By providing objective visual records of police-public contact, BWCs help build greater transparency and accountability. The footage obtained also serves as vital evidence in legal proceedings. As such, BWCs are the answer for fostering trust, alongside the safety and security of the public.
However, the transparency value of BWCs comes with significant data protection obligations under British law. Public release of footage must be weighed against legal duties to safeguard privacy and protect vulnerable groups.
There are three key requirements for managing and redacting BWC footage in the UK:
Redacting sensitive segments of footage to obscure identities and personal details is essential but painstaking when done manually. A single hour of CCTV video can contain over two million faces. Obscuring all faces manually would take days of tedious work by police staff.
This sensitive information requires careful redaction, a process that, to date, has been extremely labour-intensive and time-consuming. The current manual methods cause delays – or even full halts – in public access to necessary BWC footage. There is a clear need for a more efficient approach.
Here, artificial intelligence (AI) offers a solution. The adoption of automated, at-scale video data redaction tools can quickly and efficiently redact personal and sensitive data from the vast volumes of BWC video collected.
Using advanced machine learning, these tools can rapidly detect and blur faces, vehicle number plates, and other identifiers within video footage. This aids in the protection of private information such as the identities of minors or medical details.
For example, Pimloc’s Secure Redact platform uses AI to automatically detect and redact sensitive personal data in video up to 200 times faster than manual methods. The technology can selectively anonymise footage by removing identifying details only for non-relevant individuals.
This allows police to share video externally as needed while protecting privacy. The Information Commissioner’s Office (ICO) itself recommends using AI-enabled redaction tools as a proportionate measure.
Automated redaction is not a far-off technology of the future; it’s already here and being used by police forces to efficiently manage BWC footage disclosure.
The National Police Chief’s Council (NPCC) recognises redaction as a critical dependency; its Digital Policing Portfolio, launched in 2020, calls for a nationally scalable solution for the redaction of sensitive material. This strategic focus has led to the exploration of powerful technology solutions and many forces across the UK testing and deploying AI-powered redaction for video and audio.
It’s an opportune moment for wider adoption across British law enforcement to showcase how today’s advanced technology can foster deeper community trust with intelligent privacy protections. Responsible use of AI in this way demonstrates policing evolving in step with technology to better serve society’s needs.
Here again, AI-powered redaction presents a means of selectively applying facial matching only to individuals of interest, while upholding everyone else’s right to privacy.
With the right safeguards, FRT combined with redaction could identify suspects efficiently without collateral intrusion on the public. This smart compromise shows how technology can enhance operations while protecting citizens’ rights.
In summary, the effectiveness and positive results of BWCs in British policing are clear, driving their accelerated roll-out. But the growing volume of sensitive footage they produce is building public pressure for robust privacy protections. The UK has an opportunity to lead by example, pioneering privacy-preserving use of AI that sets the global standard for ethical technology in the service of justice.
