The UK Diesel Emissions Litigation — and the Architecture That Tested Vehicles Under Controlled Conditions While the Public Breathed Their Ordinary Behaviour

The Central Principles:

• A manufacturer cannot hide behind the limits of a test when it possesses knowledge beyond the test.
• The body that knows the product most deeply carries a duty to identify what the regulator failed to ask.
• A tool intended for life must be tested within life, not only within a controlled imitation of it.
• Regulatory approval distributes responsibility across the manufacturer, testing authority, regulator and government.
• Foreseeable harm creates a duty to prevent before the number of victims becomes politically impressive.
• Every permitted source of environmental harm should carry an enforceable duty to reduce and counterbalance that harm.
• When the test passes but the public still breathes the failure, the system has measured compliance without measuring responsibility.

The diesel emissions litigation raises a question that should have been asked before millions of vehicles reached the road:

If a vehicle’s emissions performance changes according to temperature, speed, engine load, altitude, driving pattern and operating conditions, then was the vehicle tested across the conditions in which people would actually use it?

Or was it tested only across the conditions the official procedure required?

Those are not the same thing.

The answer, historically, is that many of the vehicles at the centre of this litigation were approved through prescribed laboratory test cycles, not through exhaustive testing of every condition they might encounter on the road. The High Court’s July 2026 judgment concerned sample vehicles manufactured between 2012 and 2017 under the Euro 5 and Euro 6 regimes. The parties agreed that those vehicles were not legally required to remain within the laboratory emissions limit under every condition outside the type-approval test. The court also said that “normal driving conditions” did not mean virtually every condition that could reasonably occur anywhere in Europe, but conditions commonly and regularly encountered during ordinary use.

That reveals the first structural weakness.

The legal system created a specific testing frame.

Manufacturers built vehicles capable of passing that frame.

The vehicle then entered a much wider world.

The Test Did Not Reproduce the Full Life of the Vehicle

The older New European Driving Cycle, known as the NEDC, was a controlled laboratory procedure. It used a predetermined driving cycle rather than following one vehicle across a full year of weather, traffic, road gradients, driving styles, engine loads, maintenance conditions and ordinary human use.

Later, the Worldwide Harmonised Light Vehicle Test Procedure replaced the NEDC for new cars. WLTP was designed to resemble real driving more closely, using low, medium, high and extra-high phases, but it remained a laboratory test carried out under specific conditions. The Vehicle Certification Agency itself says that WLTP provides a closer representation of reality while warning that actual results still vary with driving style, vehicle accessories and additional weight.

Real Driving Emissions testing was introduced because governments already knew that the laboratory was not telling the whole story. From September 2017, new models had to undergo a roughly 90-minute road test combining urban, rural and motorway driving. The reform was explicitly presented as a response to the Volkswagen scandal and as a means of reducing the gap between test performance and real-world emissions.

The government’s own earlier emissions-testing programme had found a large difference between regulated laboratory NOx measurements and performance on the road. It recognised that real traffic, driver behaviour and ambient temperature affected emissions and that on-road testing was crucial for understanding the gap.

So the answer to the foundational question is clear:

The vehicles were not initially required to prove the same emissions performance across every condition in which they could realistically be driven.

They were required to pass the test that existed.

But passing the assigned test does not settle whether the manufacturers fulfilled the whole of their responsibility.

Compliance With the Test Is Not the Same as Responsibility for the Product

Manufacturers may say that they followed the procedure provided to them.

They submitted the vehicle.

They completed the required cycle.

They achieved the required result.

They obtained approval.

But a manufacturer possesses knowledge the testing authority may not possess.

It knows how the engine-control system has been calibrated.

It knows which temperatures change the operation of exhaust-gas recirculation.

It knows when emissions controls are reduced.

It knows what happens under heavier loads, unusual speeds, colder weather, hotter weather, repeated short journeys and ordinary road conditions.

It knows where the laboratory representation ends and the vehicle’s wider behaviour begins.

That knowledge creates responsibility.

If a company knows that performance changes substantially beyond the official test boundary, it should not be enough to say:

We answered only the question the test asked.

The company also carries a duty to identify the questions the test failed to ask.

A regulated industry should not treat the regulator’s limited knowledge as commercial permission to remain silent about its own greater knowledge.

If the manufacturer understands that emissions increase under conditions common enough to form part of ordinary vehicle life, then the responsible action is to disclose that variation, expand testing, request clarification or demonstrate why the change is technically necessary.

Following an incomplete standard does not make the company complete.

The Manufacturer Should Not Benefit From Knowing More Than the Regulator

The legal relationship becomes inverted when the organisation with the most technical information is permitted to do only what the regulator knows to request.

Regulators create general tests.

Manufacturers create the individual machines.

The manufacturer therefore knows the product more intimately than the authority approving it.

That informational advantage should increase its duty of disclosure.

It should not become a loophole.

The governing emissions regulation required manufacturers to design, construct and assemble vehicles so that components affecting emissions allowed the vehicle to comply in normal use. It also prohibited devices that reduced the effectiveness of emissions-control systems, subject to limited exceptions such as immediate engine-safety needs.

The law therefore did not say only:

Pass one laboratory cycle and forget everything beyond it.

It also created obligations around normal use, durability, production conformity and in-service performance.

Yet the exact relationship between laboratory limits, normal use and defeat-device definitions became legally contested. The 2026 High Court adopted a narrower interpretation, requiring proof that a device sensed test-related parameters and operated for the purpose of making emissions controls work more effectively during the test than outside it. Merely showing reduced effectiveness beyond the test conditions was not enough under that approach.

This means the manufacturers’ legal victory on most allegations does not answer the wider responsibility question.

It tells us what the court concluded the prohibition covered.

It does not establish that the testing architecture captured the full environmental reality of the vehicles.

The Law Tested Intention While the Public Experienced Outcome

The court focused heavily on whether the technology had the required relationship to the test itself.

Did it sense the test?

Was it designed to behave differently because it recognised testing conditions?

Was the purpose to improve performance during scrutiny?

Those are legitimate legal questions because intention and statutory definition matter.

But the public-health question is different:

What came out of the exhaust during ordinary use?

A person breathing polluted air does not breathe the manufacturer’s legal intention.

They breathe the output.

A child living beside a busy road does not experience whether a temperature-based calibration was designed to cheat a test or to protect an engine.

They experience the air.

A legal system may need to distinguish deliberate test manipulation from technically justified variation.

But environmental and public-health responsibility cannot disappear merely because a harmful outcome does not fit the narrowest definition of a prohibited device.

If the emissions rose greatly under ordinary conditions, that remains information requiring correction, prevention and public accountability.

The court itself acknowledged that testing the effect of individual calibrations was difficult and that none of the available approaches to isolating their NOx impact was entirely satisfactory.

That uncertainty should not become a resting place.

It should become an instruction to improve the testing system.

A Product Used in Life Must Be Tested in Life

You cannot completely test street life inside a laboratory.

A laboratory can isolate variables.

It can produce repeatable comparisons.

It can identify technical behaviour under controlled conditions.

That has value.

But the laboratory is not the final stage of the product.

The road is.

The car will not spend its life attached to a dynamometer under a prescribed temperature while following a repeated test cycle.

It will move through:

• winter and summer;
• city congestion;
• rural roads;
• hills and flat ground;
• short trips and long journeys;
• different drivers;
• different loads;
• stop-start traffic;
• motorways;
• maintenance deterioration;
• ageing parts;
• cold starts;
• heatwaves;
• humidity;
• and years of ordinary use.

A responsible approval structure must test both control and reality.

The laboratory asks:

Can we reproduce and compare the result?

The road asks:

What does the product actually do while living among people?

Neither should replace the other.

The problem begins when the controlled result is treated as though it is the whole truth.

One Vehicle Should Live on the Road Before Millions Are Released

Before a new vehicle model is sold at scale, representative vehicles should be placed into monitored real-world use for an extended period.

Not one carefully prepared afternoon.

Not one predictable route.

Not only one season.

A sufficiently long trial should expose the model to ordinary variation.

A year would allow testing through changing weather, temperature ranges, traffic patterns and vehicle ageing. It would create a fuller picture of what the model produces when it is not only being watched under the precise conditions for which it was prepared.

The data should include:

• laboratory emissions;
• real-world emissions;
• cold and warm starts;
• low and high temperatures;
• urban, rural and motorway driving;
• different loads;
• altitude and gradients;
• prolonged congestion;
• component deterioration;
• fuel and maintenance variation;
• changes caused by software updates;
• and performance throughout the full testing period.

The point would not be to prove that every individual vehicle will behave identically.

It would be to discover the behavioural range of the model before the public becomes the test population.

Commercial release should not be the beginning of foundational safety discovery.

The Public Should Not Become the Unconsenting Trial

This failure appears across industries.

Technology is released before its social consequences are sufficiently understood.

Platforms scale before psychological dependency is addressed.

Artificial intelligence reaches the public before clear responsibility, safety boundaries and escalation structures exist.

Products are normalised first.

The inquiry comes after harm.

The government repeatedly allows the population to become the real-world experiment.

Then, once enough people are affected, it calls the outcome unforeseeable.

But many harms are foreseeable when the right experts are gathered before deployment.

Psychologists can identify attachment and compulsion risks.

Sociologists can examine changes in social behaviour.

Engineers can identify technical failure paths.

Doctors can evaluate health consequences.

Environmental scientists can model cumulative exposure.

Ethicists can identify conflicts of duty.

Lawyers can locate responsibility gaps.

People with lived experience can identify how the product may actually be used rather than how its creators imagine it will be used.

Government should bring those forms of knowledge together before scale.

It should have a prevention body whose purpose is not to wait for harm, but to ask what a new system could become once placed inside real life.

Government Needs a Prevention Team and a Feedback Team

A prevention team would examine emerging products, systems and industries before mass deployment.

Its questions would include:

What changes when this reaches millions of people?

What human dependency could form?

What environmental burden could accumulate?

What groups carry the greatest risk?

What failures become irreversible?

What does the company know that the regulator does not?

What happens outside controlled use?

What compensating infrastructure is required?

A feedback team would then gather the experience of the public after deployment.

Its purpose would not be merely to close complaints.

It would connect them.

One driver reports unusual emissions behaviour.

One engineer reports a calibration discrepancy.

One researcher identifies a laboratory-road gap.

One community reports poor air quality.

One regulator receives unexplained data.

These should not remain separate files.

The system should be capable of recognising that different voices may be describing one architecture.

Prevention asks what could happen.

Feedback identifies what is beginning to happen.

Together, they create living oversight.

The Government Also Knew the Test Was Incomplete

Responsibility does not rest with vehicle manufacturers alone.

Government and regulatory bodies knew laboratory and real-world performance could diverge.

They had access to transport research.

I’d add a level of negligence and ignorance if they are drivers themselves, to the degrees of their responsibility.

They understood that road conditions vary.

They knew that an engine’s performance changes with temperature and load.

They knew the public would use cars outside the laboratory.

The introduction of Real Driving Emissions testing itself demonstrates that the need for real-world measurement was recognised.

The question is why this recognition came after years of vehicles being approved under a narrower system.

Why did the architecture wait for scandal before expanding the test?

Why was the legal standard shaped around controlled performance while public-health consequences were dispersed through ordinary use?

Why did manufacturers receive approval through conditions that did not fully represent the environment in which their product would operate?

Why did the government allow the testing gap to persist when cars were already known to be major sources of urban pollution?

The regulator cannot say only:

The manufacturers passed our test.

The regulator must also answer:

Why was that our test?

Approval Distributes Responsibility

When government approves a vehicle, it creates public trust.

The ordinary buyer cannot inspect software calibration.

They cannot measure NOx at different temperatures.

They cannot determine whether engine protection genuinely requires reduced emissions control.

They rely upon certification.

They rely upon the fact that someone with authority and technical access has already investigated the product.

Approval therefore cannot operate as a shield for the approving body.

It is an assumption of responsibility.

The manufacturer remains responsible for the vehicle it created.

The testing authority remains responsible for the adequacy of the test.

The government remains responsible for the legal framework.

The dealer remains responsible for representations made to buyers.

The board remains responsible for governance.

The industry remains responsible for what it normalised.

The approval mark does not eliminate liability.

It maps the people who participated in making the product publicly trustworthy.

Environmental Responsibility Cannot Begin and End With Punishment

If government truly valued environmental and public health, it would not focus only on whether one manufacturer crossed one legal boundary.

It would build the conditions that reduce harm across the whole system.

It would ask:

Cars now exist at scale. What must be built around them to reduce their consequences?

Industry produces emissions. What must be required to counterbalance them?

Cities retain heat. Where is the living infrastructure?

Roads dominate public space. Where are the trees, vegetation and cooling systems?

Transport remains necessary. Where is the investment in cleaner propulsion, public transport and reduced-emission mobility?

The purpose should not be to wait for pollution and then identify one offender.

It should be to design a society in which foreseeable pollution is continuously reduced.

Counterbalance Should Be Built Into Permission

Every permitted harmful activity should carry a corresponding counterbalancing duty.

If a company is allowed to produce emissions, it should contribute proportionately to reducing the environmental burden it creates.

If vehicles are allowed to enter cities, the transport system should be accompanied by serious investment in cleaner technology, public transport, active travel and urban greening.

If buildings replace natural ground, planning should require living infrastructure capable of reducing heat and supporting air quality.

If an industry consumes environmental capacity, restoration should be part of the cost of operating.

Permission should not mean:

You are allowed to create the harm.

It should mean:

You may proceed only within a structure that prevents, limits and counterbalances the foreseeable harm.

That is the difference between regulation as paperwork and regulation as stewardship.

The Paul Powlesland Contradiction Returns

This is where the case reconnects with the investigation into Paul Powlesland and the River Roding cleanup.

The legal system can investigate a person who physically removes accumulated human waste from a river because the correct permission may not have been obtained.

At the same time, large commercial systems can produce pollution through approved structures for years because they complied with the authorised process.

That contradiction exposes what the state is really prioritising.

Is it prioritising the environmental result?

Or is it prioritising its authority over who is permitted to produce the result?

If the priority is environmental health, then a restored habitat should matter.

Cleaner air should matter.

Reduced pollution should matter.

Real-world outcomes should matter.

If the priority is procedural validation, then the central question becomes whether the government authorised the action—even where authorised conduct creates more harm than unauthorised restoration.

That is a system seeking validation of its control rather than fulfilment of its purpose.

The permit becomes more important than the river.

The test becomes more important than the air.

The procedure becomes more important than the life it was supposed to protect.

A Government Investing in Destruction Weakens Its Claim to Environmental Authority

A government cannot present itself as the highest moral guardian of environmental and public health while directing public abundance toward systems of destruction and neglecting prevention at home.

It cannot fund weapons, conflict and environmentally destructive activity while treating environmental responsibility as something ordinary people and individual companies must carry alone.

It cannot criminalise restorative initiative while failing to build the infrastructure that would make such initiative unnecessary.

This does not mean companies should escape scrutiny because government also fails.

It means government must be placed inside the same responsibility architecture.

The government should not judge environmental conduct from a position outside the system.

It funds.

It approves.

It regulates.

It plans cities.

It controls infrastructure.

It sets research priorities.

It determines what becomes commercially viable.

It decides how much living space survives between roads, buildings and concrete.

It therefore participates in the conditions producing the harm.

Cities Should Not Be Built as Heat and Pollution Chambers

A city dominated by brick, asphalt, concrete, vehicles and limited canopy cannot behave like a living environment.

During heatwaves, hard surfaces absorb and retain heat.

Shade becomes scarce.

Air movement can be restricted by the built form.

People experience the accumulated result through discomfort, respiratory pressure and increased heat exposure.

Trees and vegetation can support shade, cooling, biodiversity and aspects of air-quality improvement. But they should be planned properly and cannot substitute for reducing emissions at their source. In some tightly enclosed streets, poorly designed planting can even affect pollutant dispersion. The wider principle remains: nature-based infrastructure should be deliberately integrated alongside direct emissions reduction, not offered as decoration after urban systems have already been built around heat and pollution.

Parks alone are not enough when the routes between them remain exposed.

Nature cannot be confined to designated squares while the rest of civic life is paved over.

Living infrastructure should follow people through the city.

Along roads.

Around schools.

Beside housing.

Across transport routes.

Within new developments.

Where pollution and heat actually accumulate.

Government Should Invest in the Alternatives Before Demanding Sacrifice

The public is often told to drive less, consume differently and absorb the cost of transition.

But government possesses the scale of funding and coordination required to make alternatives possible.

It can invest in:

• reliable public transport;
• electric and low-emission infrastructure;
• hydrogen research where technically appropriate;
• advanced battery development;
• cleaner industrial processes;
• urban greening;
• safe walking and cycling routes;
• freight reform;
• emissions monitoring;
• long-term environmental research;
• and new propulsion technologies capable of reducing dependence on combustion.

Not every speculative technology will prove viable. That is exactly why serious public research is required.

The state should not wait for private profitability to determine which life-supporting innovation deserves exploration.

Its role is to invest where the public need is clear but the commercial return may be too slow, uncertain or widely distributed for private actors to lead alone.

Foreseeability Creates Responsibility Before Certainty Arrives

Governments frequently behave as though they need proof of widespread harm before acting.

But prevention operates earlier than proof.

Foreseeability does not require knowing every future victim.

It requires recognising the credible pathway.

Cars burn fuel.

Combustion produces pollution.

Vehicles will be used under varying conditions.

Laboratory performance will not perfectly reproduce street behaviour.

Urban populations will experience cumulative emissions.

Those propositions were sufficient to justify broader testing and counterbalancing infrastructure before millions of vehicles entered circulation.

The same principle applies to AI and social media.

You do not need to know the identity of the person who may become addicted, manipulated or psychologically harmed.

You need to understand how human attention, attachment, reward and social reinforcement operate.

A prevention system does not wait for the final number.

It acts upon the credible direction.

The Employer–Government Contradiction

Government says that it works for the public.

But in an ordinary working relationship, the employer possesses the information required to direct the employee.

The public, however, is regularly told that government cannot disclose the full picture.

Officials know where the money goes.

Officials know which negotiations occurred.

Officials know what companies submitted.

Officials know what risks were identified.

Officials know what was excluded from public view.

The people supposedly being served know less than the institution claiming to work for them.

That reverses the relationship.

If government works for the public, information should flow toward the public.

The public should not be placed beneath the employee in knowledge, authority and control while remaining above it only rhetorically.

The lack of transparency reveals that the relationship is not structured as service.

It is structured as guardianship without sufficient public visibility.

Nothing Disharmonious Should Be Treated as Final

Many of these systems have survived because people were taught that this is simply how things work.

Cars are tested this way.

Government operates this way.

Industries regulate themselves this way.

Public bodies respond this way.

Laboratories approve products this way.

But longevity does not create correctness.

A structure should not be treated as final merely because it has become familiar.

Finality is appropriate where a system is harmonious enough to fulfil its purpose without transferring hidden harm onto others.

Where disharmony remains, the structure is unfinished.

Legislation should therefore remain open to expansion.

Testing standards should remain open to reality.

Approval systems should remain open to feedback.

Government should remain open to correction.

The law should not preserve an incomplete answer merely because it once became official.

Curiosity Must Be Directed Toward the Thing Being Judged

A proper investigation requires direct engagement.

You cannot fully understand a person only through what outsiders say about them.

You cannot fully understand a vehicle only through a laboratory representation.

You cannot fully understand an industry only through its compliance documents.

You cannot fully understand a public system only through its official report.

You must engage the thing itself.

Drive the car.

Measure the emissions.

Observe it across conditions.

Speak to the engineers.

Speak to the drivers.

Speak to the communities breathing the air.

Examine what the regulator was told.

Examine what the company knew.

Examine what happened when the product left the laboratory.

Curiosity is not the absence of judgment.

It is what makes judgment worthy of authority.

The Central Questions

The diesel emissions litigation should therefore ask more than whether the software met one legal definition.

It should ask:

Did the official test represent the conditions in which the vehicle would actually live?

What did the manufacturers know about performance outside that test?

Did they disclose the full behavioural range of the technology?

Did regulators possess enough information to approve the product responsibly?

Why were real-world tests not foundational from the beginning?

What preventive infrastructure accompanied the decision to permit millions of emitting vehicles?

What environmental counterbalances were funded?

Who monitored the lived result?

Who received the first warning?

Who failed to expand the test?

Who benefited from the gap?

Who breathed it?

The Central Principles

A manufacturer cannot hide behind the limits of a test when it possesses knowledge beyond the test.

The body that knows the product most deeply carries a duty to identify what the regulator failed to ask.

A tool intended for life must be tested within life, not only within a controlled imitation of it.

Regulatory approval distributes responsibility across the manufacturer, testing authority, regulator and government.

Foreseeable harm creates a duty to prevent before the number of victims becomes politically impressive.

Every permitted source of environmental harm should carry an enforceable duty to reduce and counterbalance that harm.

When the test passes but the public still breathes the failure, the system has measured compliance without measuring responsibility.