10 October 2026

10 October 2026: When Did Councils Become Surveillance Agencies?

Seen. Not Processed. | Investigation Update

Local councils are responsible for collecting our rubbish, maintaining parks, managing housing, providing local services and helping to keep our communities safe.

They also have to make difficult decisions about how to spend public money.

Which makes a recent development particularly interesting.

Why are councils investing millions of pounds in technology capable of identifying individuals and tracking their movements across entire towns and boroughs?

And, perhaps more importantly, why do they need to operate such capabilities themselves?

Millions of pounds — but what are we actually buying?

In September 2025, Hammersmith & Fulham Council approved £3.2 million of investment in CCTV and artificial intelligence technology.

Its subsequent council minutes describe plans for live facial recognition at ten locations, retrospective facial recognition across its existing network, automatic vehicle tracking and upgrades to 500 cameras for enhanced video searching.

That is on top of an existing network of more than 2,000 cameras and a previous £5.4 million CCTV programme.

Meanwhile, Harlow Council announced nearly £500,000 of investment in June 2026 for a new integrated surveillance network.

Its proposed technology includes facial recognition, Automatic Number Plate Recognition (ANPR), intelligent mapping and AI-assisted monitoring.

Both investments cover much more than facial recognition alone, and some of the promised capabilities are not yet confirmed operational.

Nevertheless, these are substantial commitments of public money.

At a time when councils must balance competing demands for housing, public facilities, essential services and community support, such investments deserve proper scrutiny.

We are not suggesting that every council involved is facing a financial crisis.

Even a financially secure council should have to demonstrate why a particular expenditure represents good value for its residents.

And the question is not simply whether modern CCTV can help prevent crime.

It plainly can.

The question is what additional public benefit justifies spending money on facial recognition and automated tracking capabilities that conventional CCTV does not provide.

Why does a council need facial recognition and vehicle tracking?

Let's consider what these technologies do.

Facial recognition can compare someone's biometric characteristics with a database of identities.

ANPR records vehicle registration numbers, usually accompanied by the time and location at which a vehicle was observed.

AI-assisted video analysis can help locate a person or vehicle across recordings from different cameras.

These are powerful capabilities individually.

Combine them in an interconnected system and the potential exists to reconstruct movements through public spaces, particularly where information can be associated with an identifiable individual.

We have not established that the councils discussed here are currently linking facial recognition identities directly to ANPR records. Having both functions within a system does not, by itself, demonstrate that such searches are enabled.

But the capability and its proposed uses deserve examination.

Harlow's own announcement describes integrating facial recognition, ANPR, intelligent mapping and real-time information into one system.

Hammersmith & Fulham explicitly proposes retrospective facial recognition and automatic vehicle tracking.

What legitimate council function requires these capabilities to operate together?

Parking enforcement may justify using ANPR in defined circumstances.

Protecting a council building may justify conventional CCTV.

Investigating serious crime may justify the police conducting targeted enquiries using CCTV footage.

But none of those objectives automatically explains why a local authority needs broad biometric identification and automated movement-tracking capabilities.

Where is the assessment identifying that need?

What less intrusive alternatives were considered?

And what prevents a system introduced for serious crime investigations from eventually being used for increasingly routine matters?

Those are questions of necessity and proportionality, not opposition to public safety.

Isn't this what the police are for?

Local authorities do have statutory community-safety responsibilities under the Crime and Disorder Act 1998.

They work with police, share relevant information and contribute to preventing crime and antisocial behaviour.

There is nothing inherently inappropriate about that relationship.

But supporting the police is not the same as carrying out policing functions independently.

Hammersmith & Fulham says its new technology will support the Metropolitan Police alongside the council's own Law Enforcement Team.

Harlow describes providing powerful new tools to the council, police and other enforcement agencies.

That brings us to an important question.

Who will actually operate these systems, and who authorises their use?

A police officer is subject to an established policing command structure, operational policies and professional standards.

Council CCTV operators and enforcement staff also have legal obligations, training requirements and management oversight. Civilian operation is not inherently unlawful or inappropriate.

But the two accountability structures are not automatically identical.

If a council employee can use facial recognition to search for someone across hundreds of cameras, what authorisation is required?

Can council enforcement officers initiate such searches, or must they come from the police?

Who decides which individuals may be searched for?

Who checks whether those searches were justified?

And who has authority to associate facial-recognition results with other information, including vehicle movements?

We need to establish whether these powers are properly separated and controlled, rather than assume that purchasing sophisticated surveillance software automatically confers an appropriate reason to use every function.

Counterterrorism is not a complete explanation

Broad claims about serious crime and counterterrorism deserve examination too.

Of course local authorities should help the police respond to terrorism and serious threats.

But invoking those objectives does not, by itself, explain why a council needs to acquire and operate advanced biometric surveillance capabilities.

If the police need to identify and follow a suspect during a serious investigation, why isn't conventional CCTV evidence supplied to them under established procedures sufficient?

What additional capability does the council itself need?

Who will use it?

Under what conditions?

And how frequently is it expected to be required?

A genuine operational requirement should be capable of being explained and documented.

Public safety is an objective. It is not a blank cheque for surveillance technology.

What does the public get for its money?

This is where the financial argument becomes particularly relevant.

When a council invests millions in surveillance infrastructure, the cost is not limited to cameras and software.

There may also be ongoing licensing, maintenance, training, staffing, storage, cybersecurity, technical assurance and eventual replacement costs.

So what is the complete cost of ownership?

What outcomes were forecast when the investment was approved?

How will the council measure whether those outcomes have been delivered?

And will the public be shown that assessment?

For example, if facial recognition is introduced to help prevent crime, the council should be able to identify how it intends to measure the additional benefit attributable to facial recognition, rather than improvements from CCTV generally.

The same applies to vehicle tracking and AI-assisted searching.

Faster searches and more capable technology may help investigators. That is a potential benefit worth assessing.

But improved capability is not, on its own, proof of improved public safety.

Buying something because it can do more is not the same as demonstrating that the additional capability is needed.

That principle should be familiar to anybody responsible for spending public money.

Who asked the residents?

There is also a democratic question.

Local authorities exist to serve their communities.

Their CCTV networks monitor the streets, parks, shopping areas and public buildings used by those communities every day.

Residents may support additional measures to address crime and antisocial behaviour.

But support for better public safety should not automatically be interpreted as support for increasingly sophisticated biometric surveillance.

Were residents specifically consulted about facial recognition and automated movement tracking?

Were they told what information the systems could collect and combine?

Were the alternatives and privacy implications explained?

And did councillors consider whether the public would regard these capabilities as proportionate?

We are not asserting that none of these consultations or assessments took place. We want to establish what was recorded and what was made public.

There is also a simple practical reality.

People often cannot avoid travelling through their own town centre, walking past a council camera or visiting a public-service building.

That makes meaningful transparency particularly important.

The safeguards should exist before the system is switched on

The government's Surveillance Camera Code of Practice guidance is explicit that surveillance must address an identified need, be proportionate, consider privacy and have effective governance.

Its August 2026 guidance for local authorities even warns about enabling recognition technology on existing cameras without an appropriate operational requirement or Data Protection Impact Assessment.

The Information Commissioner's Office similarly expects organisations using facial recognition to explain why the technology is necessary, why less intrusive alternatives were ruled out and how its effectiveness is measured.

Those are not unreasonable demands from privacy campaigners.

They are established expectations of public authorities.

And they bring us back to our original investigation.

What happens to biometric information generated when someone does not match a watchlist?

How is deletion verified?

Are searches audited?

Can inappropriate searches be identified?

Who has access to the information?

And what prevents information collected for one purpose from being used for something else?

The more capable and interconnected the system becomes, the more important these questions are.

Our next investigation

Seen. Not Processed. has already submitted Freedom of Information requests to North Tyneside and Preston councils examining their use of facial recognition.

We will now extend our attention to the larger AI surveillance programmes being proposed elsewhere.

For Hammersmith & Fulham and Harlow, the next questions will concentrate on the documented business case, full costs, permitted uses, operational responsibility, links between surveillance systems and the safeguards governing those capabilities.

We will distinguish between systems that have been approved, systems being installed and biometric functions confirmed as operational.

And we will publish the evidence whether it demonstrates effective safeguards or identifies gaps in accountability.

There may be sound reasons for particular deployments.

If so, the supporting documents should explain them.

But local authorities should not acquire increasingly powerful surveillance capabilities merely because the technology is available, suppliers promote its benefits or a general public-safety objective sounds persuasive.

The more intrusive the capability, the stronger the justification should be.

Councils have many important responsibilities.

Protecting their residents is certainly one of them.

But so is protecting their residents' privacy, spending their money responsibly and being accountable for the decisions they make.

Before councils invest millions in technology capable of identifying people and tracing movements across their communities, they should be able to answer one straightforward question:

Why do you need this power, and why should the public trust you with it?

Trust requires evidence.

SEEN. NOT PROCESSED.

The right to be seen, but not processed.