£81 Billion To Fraud. £28 Billion To Criminals And Enemies.

There is a number the British public should be far more aware of than it is. The National Audit Office, using Public Sector Fraud Authority quango methodology, estimates the taxpayer lost between £55 billion and £81 billion to fraud and error in 2023–24. Not across a decade. Not cumulatively. In one year.

Most of it went undetected. Most of it will never be recovered. And nobody has been fired.

Total managed expenditure for 2023–24 was approximately £1,189 billion. The NAO's estimate covers losses across both public spending and income systems, so the comparison is not perfectly direct, but the scale remains staggering: the upper figure exceeds the entire defence budget and dwarfs annual foreign aid spending so comprehensively one wonders why the aid budget attracts more controversy than the fraud it apparently funds.

The government's response has been to announce it saved £7.5 billion in the past year through counter-fraud work. This is rather like a man whose house is stripped bare every night boasting he once remembered to close a window.

Source: NAO report

These figures measure different things across different periods. They should not be lazily added together. But they should be understood as different windows into the same institutional condition.

The Anatomy of Institutional Haemorrhage

The NAO's breakdown is more damaging than its headline figure. Of the estimated losses, approximately £12 billion was detected by quangos, meaning someone noticed, flagged it, and could in theory do something about it. A further £41 billion was estimated but not detected: the authorities are fairly certain the money was lost but cannot identify the specific recipients or recover the funds.

Then comes the most revealing category of all: between £3 billion and £28 billion of potential loss sits in areas where the government does not even bother to measure fraud and error. Around £560 billion of public expenditure is simply not subject to systematic fraud assessment.

The state does not know how much is being stolen from it because it has decided, across more than half a trillion pounds of annual spending, not to look.

This is not negligence. Negligence implies someone tried and failed. This is policy.

A Counter-Fraud Apparatus Built Backwards

The government will point to its counter-fraud efforts. It always does. But the NAO's own analysis of the counter-fraud workforce reveals an apparatus built in the wrong direction: too much effort applied after payment, after loss, after dispersal, after the money has vanished into networks, accounts, shell companies and jurisdictions it will never be retrieved from.

Two-thirds of counter-fraud staff specialise in investigation rather than measurement and prevention. Officials in some public bodies told the NAO they lacked the skills even to measure fraud, let alone stop it.

This explains the government's favourite statistic. "We saved £7.5 billion" sounds impressive until you realise the savings come from chasing money already gone, not from preventing it leaving. Britain investigates the stable door after the horse has been sold, exported and given a new passport.

37 Years Of Failure In One Department

If any private company had its accounts qualified by auditors for thirty-seven consecutive years, its directors would be disqualified, its shareholders would revolt, and the nutjob Financial Conduct Authority would send letters of a deeply unfriendly character. If it happened for a decade the press would call it a scandal. If it happened for two decades the company would cease to exist.

The Department for Work and Pensions has achieved precisely this distinction and continues to receive its budget on time.

For thirty-seven successive years, the Comptroller and Auditor General has qualified DWP's accounts due to the material level of fraud and error in benefit expenditure. The Public Accounts Committee has called the streak "unacceptable" and noted, with a hint of gallows humour, the department is three years away from a "sad and embarrassing milestone" of four full decades of audited failure.

Stop here.

37 YEARS of unauditable accounts, because of fraud levels.

In 2023–24, DWP overpaid £9.7 billion in benefits. Of this, £7.4 billion was attributed to fraud. Universal Credit alone had an overpayment rate of 12.4 per cent, meaning roughly one pound in every eight disbursed through the system was paid to the wrong person or in the wrong amount. The fraud component of Universal Credit spending was 10.9 per cent, or £5.66 billion. For context, £5.66 billion is a third of the entire budget of the Foreign Office.

DWP's stated ambition is to reduce the overpayment rate to 2.8 per cent by 2028–29. It has described achieving this as "impressive." The PAC was not persuaded. Nor should anyone else be. "Impressive" is what you call a dog balancing a biscuit on its nose, not the target for a department disbursing a quarter of a trillion pounds per year.

Meanwhile, 143,922 unpaid carers were, as of early 2025, carrying Carer's Allowance overpayment debts totalling £250 million, much of it accumulated because DWP took years to act on its own earnings-breach alerts. Carers who had breached the weekly threshold by a few pounds found themselves issued demands for thousands. The number of affected carers rose 71 per cent between 2018–19 and 2023–24, not because carers became dishonest but because the department could not be bothered to send a letter.

This is the moral obscenity of the system at Lego level: it is simultaneously too weak against fraudsters and too brutal against the honest. It loses billions to organised criminals and then sends a debt collector to a woman caring for her disabled husband because she earned £3 above a threshold nobody told her had changed.

The Covid Fraud Machine

Covid relief exposed the outgoing payment system. Procurement exposed the purchasing system. Both failed for the same reasons.

The Covid Counter-Fraud Commissioner's independent report found £10.9 billion was lost to fraud and error across pandemic relief schemes. The government's own announcement conceded the schemes were rolled out with "huge fraud risks and no early safeguards."

The Bounce Back Loan scheme alone generated £4.1 billion of fraud and error.

There was at least one reported case of a recipient channelling pandemic loan funds to Islamic State in Syria.

On the purchasing side, the Department of Health and Social Care wrote down approximately £15 billion on pandemic PPE over two financial years, including items bought at vastly inflated prices, equipment deemed unfit for use, stock left to expire in shipping containers, and material so defective it could not even be given away. The Public Accounts Committee reported £9 billion of the £12 billion spent on PPE in 2020–21 was wasted through inflated pricing or shoddy equipment.

Two commercial waste companies were appointed to incinerate 15,000 pallets of unusable PPE per month.

Covid did not create these weaknesses. It merely applied pressure to a machine already riddled with them. The same credulity, the same absence of verification, the same institutional reluctance to say "no" at the point of payment existed long before anyone had heard of SARS-CoV-2.

The pandemic was a stress test. Britain failed it the way a student fails who has not revised, not because the exam was unfair but because he never once opened the book.

The R&D Scheme For Subsidising Fraud

Every fashionable government objective becomes a payment surface. Every payment surface becomes a feeding ground. This rule applies to every government in every country on Earth. It is the one invariable condition all planning must be originated around.

Take the R&D tax relief scheme, the centrepiece of the state's "innovation" agenda. In theory, this is an excellent idea. In practice, it is a magnet for dishonesty and evasion.

HMRC's own analysis concluded the error and fraud rate was 16.7 per cent of total claims in 2020–21, amounting to £1.13 billion. The SME scheme was worse, hitting a peak error rate of 25.8 per cent in 2021–22, meaning one quarter of all claims by value were incorrect or fraudulent.

Over the period from 2020–21 to 2023–24, HMRC accounts indicate more than £4 billion was paid out as a result of fraud and error.

Tax Policy Associates estimate the true figure, including years before systematic measurement began, may be closer to £10 billion. The NAO qualified HMRC's accounts on the same basis.

A scheme designed to reward ingenuity instead rewarded dishonesty. The tax advice market is unregulated. Consultants pressured companies into making fraudulent claims, sometimes without the company's knowledge. A House of Lords committee was told in 2022 of a "wild west" of R&D tax agents. The fraud rate for R&D relief was more than three times the rate across the broader tax system.

Britain built an incentive for research and created an incentive for theft. Nobody redesigned the scheme until the losses became publicly undeniable, and even then the primary government response was to open a voluntary disclosure facility asking fraudsters to come forward and confess. One can only marvel at the optimism.

The Tax Gap And Selective Blindness

HMRC's published tax gap for 2023–24 stood at £46.8 billion, representing 5.3 per cent of total theoretical tax liabilities. Not all of this is fraud. The behavioural breakdown includes failure to take reasonable care (31 per cent), error (15 per cent), evasion (14 per cent), criminal attacks (9 per cent), non-payment (12 per cent), legal interpretation (12 per cent), the hidden economy (around 5 per cent) and avoidance (1 per cent).

There is an instructive asymmetry.

The state is simultaneously overpaying the wrong people through benefits and grants and failing to collect what it is owed through tax.

It haemorrhages on both sides of the ledger: funds fraud on the way out, tolerates evasion on the way in. This is not an accounting problem. It is an institutional condition, and it has been tolerated for so long it has become structural.

A State Unable To Audit Itself

Fraud thrives where accounts are late, fragmented or unauditable. On this measure, Britain is not merely vulnerable but inviting.

In November 2024, the head of the NAO issued a disclaimed audit opinion on the Whole of Government Accounts for 2022–23, the first time in history the national books had been refused sign-off. Of England's 426 local authorities, only 43 (10 per cent) had submitted reliable audited data. Another 196 (46 per cent) submitted unaudited figures. And 187 councils (44 per cent) submitted nothing at all.

It got worse.

For 2023–24, of 407 councils due to report, only 16 (4 per cent) submitted adequate audited data. Fifty-five per cent submitted unaudited figures. Forty-one per cent again submitted nothing. The NAO refused to sign off the national accounts for the second consecutive year.

The chair of the Public Accounts Committee warned the government was "flying blind" on local authority finances and said more than two-fifths of councils had indicated they would go bankrupt within two years without emergency Treasury funding.

If the state cannot produce audited accounts for its own local-government tier, every claim it makes about fraud detection, prevention, or recovery elsewhere is built on sand. Britain is not merely being defrauded. It cannot demonstrate it is not being defrauded, because it cannot produce the accounts.

Grants To Companies Linked To Hostiles

The aid and grants picture is worse than any departmental spreadsheet has ever admitted. Officially detected fraud in UK aid is tiny. The Independent Commission for Aid Impact reported the FCDO discovered just £2.2 million of fraud in 2020–21 against £9.9 billion of spend, a detected loss rate of 0.02 per cent.

ICAI itself noted the implausibility of this figure, given the Public Sector Fraud Authority's own cross-government estimate of 0.5 to 5 per cent losses. The clean numbers were never the whole picture. They were the picture the department wished to present.

In November 2025, BBC Panorama reported on AidData research showing £45 billion of Chinese state-backed investment in UK businesses and projects over recent decades, some of which gave China access to military-grade technology. The UK was the top destination among G7 nations for these investments relative to population and economic size.

Jeremy Fleming, former head of GCHQ, stated Britain had been:

far too free in allowing access to strategically important industries.

Dynex Semiconductor, a Lincoln-based company owned by the Beijing-headquartered CRRC Corporation, received £3 million in UK government grants and support over the past decade. CRRC was placed on a Pentagon blacklist, accused of being a "Chinese military company." Apparently nobody in Whitehall checked who owned the grantee before writing the cheque.

The point of entry is not a masked man with a bag of cash. It is a company registration, a grant application, a nominee director, a subcontractor, a research proposal. Modern hostile finance does not need to storm the Treasury. It fills in the forms. And Britain, with its threadbare due diligence and its touching faith in paperwork, processes the application and sends the money.

The Buried Dossier And The Security Inversion

In June 2026, the Telegraph reported the existence of a secret Cabinet Office assessment finding more than £28 billion of taxpayer money reached actors linked to national-security threats between 2015 and 2021. The report, commissioned by security officials in 2023 (i.e. MI5, etc), was allegedly buried to spare the government political embarrassment.

The Telegraph's account is staggering.

It includes grants to companies linked to the Russian state, Covid loans reaching Islamic State-linked recipients, and research investment directed to companies connected to the Chinese military. The buried report:

... demonstrates that foreign aid and Covid relief loans were appropriated on a vast scale by Britain’s enemies, with the money beyond reach and those who took it unpunished.

More than £28bn ended up in the hands of those wishing to harm Britain between 2015 and 2021, according to the report, which was commissioned and produced by the Cabinet Office but was buried during the previous government.

A large proportion of the misappropriated funds reportedly went to criminal gangs, including human-traffickers claiming housing benefits and disability allowances. Sources described a concerted effort by an organised crime network linked to Eastern Europe and backed by an unnamed hostile state to obtain British public funds and encourage illegal immigration.

The £28 billion figure is not yet independently auditable because the dossier remains unpublished. But the claim fits every known pattern: large-scale fraud, weak due diligence, poor data, departmental fragmentation and the near-complete absence of national-security screening on grant disbursements.

If the report is wrong, publish it and show the methodology; if it is right, the public has been denied one of the most serious assessments of state failure in modern British government.

And if the report was buried because the findings were embarrassing, the civil servants who made the decision should be named, because concealing a national-security assessment for political convenience is itself a security failure.

Tom Keatinge of the RUSI think tank has long argued the benefits system has become an "ATM for terrorists." His 2021 paper described the lack of attention given to fraud in the national-security discussion as "increasingly perverse" and called for a greater role for government intelligence services.

Professor Nicholas Ryder of Cardiff University, a former adviser to the home affairs select committee, agreed the Telegraph the findings were "staggering" and the link between fraud and terrorist financing was "very clear." He added the US and Australia devoted far more resources to using fraud investigations as a mechanism to disrupt terrorism.

Fraud is no longer a fiscal offence. It is a hostile-finance channel.

When organised crime, terror networks and hostile states exploit public payment systems, fraud ceases to be a Treasury matter. It becomes an intelligence failure. Britain has had the intelligence failure. It wrote a report about it. Then it hid the report.

The Inversion Of Every State Duty

The moral structure of this failure is worth making explicit, because the people responsible will try to obscure it with announcements, targets, recovery statistics and the ritual sacrifice of one or two low-ranking fraudsters prosecuted pour encourager les autres.

The British state taxes its citizens, borrows against their children, cuts services, pleads poverty, and then loses sums large enough to fund entire departments into fraud, error, criminal networks and, reportedly, the enemies of the country.

• It scrutinises ordinary citizens but cannot scrutinise its own payments.
• It lectures taxpayers on responsibility, then misplaces their money.
• It claims compassion, then funds criminals.
• It claims security, then subsidises threats.
• It claims poverty, then leaks billions.
• It demands audited accounts from every business in the land, then cannot produce audited accounts for its own local-government tier.

The common mechanism across every failure is the same. X is prioritised politically over Y, aka common sense; yielding Z, predictable catastrophe.

• Speed over verification, in Covid relief.
• Entitlement over identity, in welfare.
• Moral mission over due diligence, in aid and grants.
• Complexity over enforcement, in tax.
• Departmental silos over national-security screening, in everything.
• Secrecy over accountability, in the buried dossier.
• Investigation over prevention, in the counter-fraud workforce itself.

These are not separate problems. They are the same institutional pathology expressed across different spending categories. The state has grown too large to police, too sentimental to distrust, too fragmented to coordinate, and too embarrassed to confess what it already knows.

An ATM For Terrorists, Fraudsters, And Enemies

Britain has built a state so large, so careless and so morally self-congratulatory it cannot distinguish citizen from fraudster, business from shell company, relief from hostile finance, or grantee from enemy.

Its payment systems have become, in the words of one think-tank analyst, an "ATM for terrorists." Its benefit machinery has been qualified by auditors for nearly four decades. Its innovation schemes were exploited by unregulated agents running a "wild west." Its pandemic response was conducted with "huge fraud risks and no early safeguards." Its procurement machine wrote down £15 billion of PPE and incinerated the evidence at 15,000 pallets a month. Its local councils cannot produce accounts. Its national books have been refused sign-off for two consecutive years. And when a secret report quantified the full security cost, the civil service hid it.

The British public has not been told how bad it is, because telling them would require admitting every claim made on their patience and their purse was made under false pretences. Every plea for higher taxes, every cut to local services, every speech about "difficult choices" was delivered by a state which was simultaneously haemorrhaging money into the hands of people who were lying to it, stealing from it, or actively working to destroy it.

A state unable to tell who it is paying cannot call itself generous. It cannot call itself competent. It cannot call itself secure. And in the deepest sense it cannot call itself sovereign, because a sovereign institution must at minimum know whether the money it disburses on behalf of its citizens reaches friend or foe.

Britain does not know. And until very recently, it preferred not to ask.