The Quote That Moved While You Slept: AI Car Insurance Pricing and Your UK Rights

You check your car insurance renewal on a Monday morning. The figure is unpleasant. By Tuesday evening, after no change to your car, address or claims history, it has moved again — sometimes up, sometimes down. It can feel like the quote is being pulled out of a hat. In reality, it is often being produced by a pricing engine that has learned from years of claims data, quote behaviour and market signals.

At AIWF, we are not here to tell you that every insurer is doing something dodgy. Most are trying to price risk and stay solvent. But when the number changes overnight, you deserve to know what is going on, what rules apply in 2026, and how to push back if the explanation does not add up.

Why the same driver gets a different number

Car insurance pricing has never been one clean sum. Even before AI became a common shorthand, insurers used statistical models to estimate risk. What has changed is the scale and speed. Modern pricing systems can weigh hundreds of signals — some about the car, some about you, some about the moment you asked for a quote.

That is why two quotes for the same driver can differ. You might have used a different comparison site, changed the voluntary excess, adjusted your annual mileage, picked a different start date, or asked for monthly payments rather than one annual sum. The insurer's appetite may also have shifted: it might be hungry for certain postcodes or drivers this week and cautious about them next week. Some of that is commercial judgment. Some of it is automated segmentation. The line between the two is not always obvious from the outside.

Renewal quotes add another layer. Since the general insurance pricing practices rules came into force in 2021, insurers have not been allowed to use loyalty penalties in the same way — the so-called price walking ban. But that does not mean renewals are identical to new-business quotes. Insurers can still reprice according to risk, costs and their own models, and the result can still feel arbitrary if nobody explains it properly.

What the model is likely looking at

Not every insurer will use every data point, and no one outside the pricing team sees the full recipe. But common ingredients include your claims history, the car's make and model, where it is kept overnight, your annual mileage, your occupation, your age and driving record, and any modifications. Payment method matters too: paying monthly is often a credit arrangement, not just a convenient split.

Some systems also look at quote behaviour — how many quotes you have run, how close you are to renewal, whether you accepted or declined add-ons. Insurers say this helps them match prices to demand and reduce fraud risk. Consumer groups worry it can drift into personalisation that is hard for customers to understand. Both things can be true.

What matters for you is the output. If a price has changed, ask which details were used and whether the insurer can explain the main reasons in plain English. You do not need to reverse-engineer the model. You do need to know whether the decision was based on accurate information.

The rules that still apply in 2026

The regulatory picture in 2026 is not a wild west. The Financial Conduct Authority's Consumer Duty requires firms to act in good faith, avoid foreseeable harm and deliver fair value. That applies whether a human underwriter or an algorithm sets the price. The FCA has been clear that existing rules do not disappear because a model is involved.

Data protection law matters too. Under the UK GDPR and the Data Protection Act 2018, you can ask for a copy of your personal data and ask how it is being used. If a decision about you is made solely by automated means and has a legal or similarly significant effect, Article 22 gives you specific protections — including, in many cases, the right to human intervention, to express your view and to contest the decision. Insurance pricing can be a complicated area here, because not every quote is a solely automated decision with significant effects. But if you are turned down, charged a materially higher premium, or treated in a way that feels final and unexplained, it is fair to ask the question.

The Financial Ombudsman Service can look at complaints if you are unhappy with how an insurer handled things. The Information Commissioner's Office can consider data protection concerns. Neither route is a magic wand, but both are real, free and worth knowing about.

A practical checklist for questioning a quote

1. Recheck the basics first. Confirm your mileage, address, overnight parking, claims history, drivers and modifications. A tiny input error can move a premium a long way.

2. Compare like with like. Use the same excess, payment method and start date across quotes. Monthly prices can look smaller while costing more overall.

3. Ask what changed. If a renewal has jumped, ask the insurer to identify the main factors. You are not demanding the source code; you are asking for a clear explanation.

4. Ask whether automation was involved. A useful question is: was this decision made solely by automated means, and can a human review it? Keep a record of the answer.

5. Make a subject access request if needed. Ask for the personal data used to calculate the quote. Insurers can redact commercial secrets, but you should still get a meaningful response about your own information.

6. Use the complaints route. Put your concern in writing. Say what you asked, what you were told and what outcome you want. If the insurer does not resolve it, ask for a deadlock letter or take it to the Financial Ombudsman Service.

7. Do not assume the comparison site is the whole market. Some insurers are not on every panel. A direct quote can sometimes beat the screen, and vice versa.

Spotting when automation is doing the talking

You will rarely see a label saying 'an algorithm decided this'. Instead, look for signs: a quote that changes without any input change, an explanation that repeats generic phrases, a chat assistant that cannot deviate from a script, or a call handler who can see the price but not the reason. None of these proves wrongdoing. They do suggest you should ask for a human review and a clearer explanation.

A calm approach works better than a furious one. Write down the date, the quote reference and the questions you asked. Ask for answers in writing. Insurers respond to clear, specific complaints because they have processes for them — and because Consumer Duty has made outcomes monitoring a board-level concern.

A calmer renewal routine

Give yourself a week before renewal. Gather your documents, run a few like-for-like quotes, and note the differences. If a price seems inexplicable, ask. If the answer is still vague, escalate. You are not being difficult. You are using the rights that already exist in 2026 — and reminding the industry that a quote is not a mystery, it is a decision that should be explained.