The weeding problem:
The past needs an audit. The future needs a switch.
Stop the automatic deletion now
24/08/2026 to the National Audit Office (NAO)
GF 3845-26 — POTENTIAL SYSTEMATIC UNDER-RECORDING OF STOLEN VEHICLES
I write further to your very helpful letter of 25 June 2026 and, in particular, the enquiries your team made of the Home Office concerning the six-week removal of unconfirmed PNC Lost/Stolen (LoS) markers.
I am contacting you directly because you explained that this matter had been passed to you as the NAO Director responsible for the Home Office, and your team had already engaged with the Home Office specifically about my concerns.
There has now been a material development which I believe justifies revisiting your conclusion that the NAO did not intend to undertake further work at that time.
Your letter recorded three particularly important findings arising from your team’s enquiries:
- an LoS marker is initially unconfirmed and only once confirmed is it included in the report available to DVLA;
- if it remains unconfirmed after six weeks, it is removed from PNC; and
- the Home Office confirmed to your team that this process causes the discrepancy between the police and DVLA databases.
You also established that the Home Office was considering whether the LoS confirmation step should be removed from LEDS, but that there was no timetable for doing so.
We now know the underlying failure is still occurring
On 24 August 2026, in response to my continuing FoIA enquiries, the Home Office confirmed that recently archived VE602J reports contain:
“details of each individual two-week reminder for lost/stolen markers, along with several other notification types.”
That appears highly significant.
The PNC procedure requires LoS records to be confirmed promptly – I understand within 48 hours (PNC User’s manual)
Yet the Home Office’s current/recently archived data contains individual LoS records which have reached two weeks without confirmation.
This is therefore not simply evidence of a historical weakness which might eventually be resolved through LEDS. The condition which can ultimately lead to six-week removal is continuing now.
That is despite:
- the issue having previously been circulated nationally by the NPCC (11/2024)
- the matter apparently having been raised nationally again subsequently;
- my own approaches to the NPCC and Home Office;
- my detailed submissions earlier this year; and
- importantly, the NAO itself raising the matter with the Home Office and asking what remedial action was proposed.
Has anybody measured the problem?
The Home Office’s latest response states that analysing its VE602J material to produce police-service-level totals would require “significant human effort”. That concerns me for a different reason.
It raises the possibility that, notwithstanding the existence of these reports and the known consequences of non-confirmation, nobody nationally is routinely analysing this information to determine the scale of the problem or whether previous remedial action has worked.
The two-week reports should potentially provide a straightforward governance measure:
- How many LoS records are failing to be confirmed within the required period, where is that occurring, and is the position improving?
The corresponding six-week information should answer the still more important question:
- How many reach the point at which their stolen marker is removed?
I respectfully suggest this now merits an audit
I fully appreciate your June position that the NAO did not intend to undertake further work at that time. But the evidential position has moved on.
The NAO established with the Home Office that the process causes a national data discrepancy. The Home Office told you that changing the process through LEDS was under consideration, without a timetable.
We now know that recent Home Office reports continue to contain individual two-week LoS reminders.
I therefore respectfully ask whether the NAO will reconsider whether this warrants an assurance or value-for-money examination.
The exercise need not begin with my statistical analysis. The existing administrative records themselves appear capable of answering the fundamental questions:
- How many two-week LoS reminders are generated?
- How many subsequently reach six weeks?
- How many LoS markers have historically been removed through that process while the vehicle remained stolen and outstanding?
And what governance has existed nationally to identify, quantify and address those failures?
There is potentially a much greater consequence. My concern has never been merely that two databases contain different numbers.
If a genuinely stolen and unrecovered vehicle loses its LoS marker, the consequences may extend considerably further.
It would be important to establish whether affected vehicles subsequently:
- acquired new registered keepers;
- were the subject of PNC enquiries;
- passed ANPR cameras;
- underwent DVLA transactions; or
- were checked through vehicle-provenance systems while no stolen marker was available.
That leads to the particularly troubling possibility that innocent purchasers may presently possess stolen vehicles which appeared not to be stolen when checked because the relevant marker had previously been removed.
I do not know how many such vehicles exist. That is precisely why I am asking for the audit.
Nor do I suggest that the entire discrepancy I have identified between police and DVLA statistics is caused by this process. I have repeatedly made that qualification.
- But your own enquiries established that the process is a cause.
The question now is its scale and consequences. But first, surely the priority must be to stop adding to the problem.
Gwent Police has previously demonstrated that six-week deletion is not inevitable: its response, following my approach, was to require LoS records to be confirmed at the point of creation. That is important because it demonstrates that the underlying automatic deletion process need not be allowed to take effect.
However, the more recent Gwent case I have identified also appears to demonstrate the weakness of relying upon manual compliance as the safeguard against an automatic process. A record apparently disappeared at approximately six weeks and was not reinstated until two months later. I am presently seeking confirmation from Gwent Police as to precisely what occurred.
If an automatic process carries the risk, demonstrated by the Home Office’s own evidence, of removing the stolen status of an unrecovered vehicle because somebody has failed to complete an administrative step, why retain that process at all?
The historic position can then be audited. But continuing to create the possibility of further affected vehicles while that exercise is undertaken seems increasingly difficult to justify.
- An automatic deletion process should not depend upon flawless manual intervention to prevent an inappropriate outcome
There is also a much simpler immediate issue. If automatic six-week removal is known to cause the discrepancy; if the Home Office is already considering eliminating the confirmation step from LEDS; and if records are demonstrably still reaching two weeks unconfirmed:
- why permit the automatic six-week removal process to continue while its scale is being established?
It appears entirely possible to separate the two exercises:
- First, prevent further potentially inappropriate removals.
- Second, audit what has already happened.
The consequences of an historic audit may be complicated. It may identify unrecovered stolen vehicles, innocent purchasers, missed investigative opportunities and difficult questions concerning responsibility. But those potential consequences surely strengthen, rather than diminish, the case for establishing the facts.
I would therefore be grateful if you would reconsider the position set out in your letter of 25 June and advise whether this new evidence warrants further NAO consideration. In particular, I respectfully ask whether the NAO would consider:
1. asking the Home Office to preserve/suspend the six-week automatic removal process pending review;
2. establishing the numbers of two-week and six-week LoS notifications nationally and by police service;
3. examining what monitoring, assurance and governance exists over those reports; and
4. considering an audit of historically removed LoS markers to establish the number relating to vehicles which remained stolen and outstanding and the consequences which followed.
If the matter now falls to another NAO director or team, I would be grateful if you would identify that person/team and arrange for this correspondence and the previous GF 3845-26 material to be passed to them.
Yours sincerely,