Eurolaser IT Limited v HMRC [2026] UKUT 00324 (TCC), published 21 August 2026, is a board-level VAT fraud control case dressed as an Upper Tribunal appeal on findings of fact. Judges Raghavan and Tilakapala dismissed the company’s appeal. The First-tier Tribunal’s decision stood: input tax denial of about £1.63m, denial of zero-rating on intra-Community supplies of about £0.50m, and penalties of about £0.31m.
The CFO problem is not “another MTIC horror story”. It is narrower and more dangerous. HMRC accepted that the sole director, Mr Pallister, neither knew nor should have known the deals were connected with fraudulent VAT evasion. The self-employed consultant who introduced and arranged the 87 purchase transactions, Mr Moshin Darr, did. The tribunal attributed Darr’s knowledge to the company. An “innocent director” did not save the input tax.
Primary sources: the Upper Tribunal decision summary and PDF, HMRC’s Kittel overview (VATF52100), the three-limb structure in VATF52300, “should have known” guidance in VATF53420, due diligence expectations in VATF73000, VAT Notice 726, and practitioner notes from Francis Wilks & Jones, InTax and VATupdate on the FTT stage.
What Kittel and Mecsek still do in 2026
The Kittel principle lets HMRC deny the right to deduct input VAT where transactions are connected with fraudulent evasion and the taxable person knew or should have known of that connection. Mecsek-type analysis sits alongside where export/zero-rating and participation in evasion are in play. Post-Brexit, the UK still applies the same core architecture through the VAT Act machinery and HMRC’s VAT Fraud manual.
HMRC’s three limbs remain the working map:
- Fraud somewhere in the extended chain — missing trader, contra-trader or related evasion, not mere commercial failure.
- Connection — direct or indirect, including contra-trading patterns.
- Knowledge or constructive knowledge — actual knowledge is rare; “should have known” asks whether the only reasonable explanation, on the objective facts, was a connection to fraud.
That third limb is where finance leaders lose money. Tribunals do not require a signed confession. They stack hallmarks: back-to-back deals, repetitive quantities, uncommercial mark-ups, high-risk goods, thin due diligence, and counterparties that do not survive basic checks. Eurolaser adds a sharper edge: the person with the bad knowledge does not have to be the director on the bank mandate.
What the Upper Tribunal actually decided
Eurolaser traded IT products. Darr, self-employed, introduced and arranged the disputed purchases. The FTT found he knew — or at least should have known — the transactions were connected with fraudulent VAT evasion, and that his knowledge was attributable to the company. It also found the company had not taken reasonable steps to prevent the fraud. The director’s own innocence was accepted by HMRC and still did not reopen input tax.
On appeal, Eurolaser ran classic Edwards v Bairstow attacks on “industry normal” mark-ups and back-to-back patterns, and on use of Darr’s earlier MTIC history with Euro Stock Shop Limited (ESSL). The Upper Tribunal refused to reopen the fact-finding. The bar is deliberately high: features that can appear in legitimate trading still matter in the cumulative picture.
On the ESSL history, the UT’s framing matters for every KYC file. Previous MTIC involvement and earlier tribunal findings were not used as pure character assassination. They were objective factors showing familiarity with how fraudulent chains operate and what the red flags look like. Someone who has lived inside MTIC patterns is harder to paint as commercially naïve when the same patterns reappear.
For CFOs, the holding in one sentence: if the person who actually sources and structures the deals knew or should have known, the company can still lose the VAT — even where the boardroom was clean on the facts of that case.
Why “we outsource procurement” is not a control
Lean head offices often park commodity buying, grey-market tech, labour, freight and wholesale with introducers, self-employed “sales consultants” or buying desks. Speed and margin are the commercial story; attribution is the tax story.
Earlier authorities already warned that agent knowledge can stick to the principal. Eurolaser is the 2026 board brief: a self-employed label is not a knowledge firewall if the consultant is the commercial brain of the purchase book.
Three failure modes show up repeatedly in diligence:
- Director comfort without counterparty file depth. “My guy knows the market” is not a risk assessment. Background on the introducer — Companies House history, director bans, prior tribunal involvement, disqualifications — belongs in onboarding, not discovery after a Kittel letter.
- Hallmarks normalised as “how this sector works”. Back-to-back, same-day chains, circular counterparties, goods that never need inspection, and margins that do not move with market risk are not automatically fatal. Cumulatively, without challenge, they are fatal.
- Due diligence as a PDF pack, not a living control. VAT number checks, credit reports and a signed terms sheet help only if someone with authority can stop the deal when the pack fails. Tribunals notice rubber stamps.
What good looks like on a finance control stack
You do not need a 40-page VAT fraud policy. You need owners, kill-switches and evidence that would survive a tribunal bundle.
1. Map who can create VAT recovery risk. Employees, contractors, introducers, buying agents, drop-ship desks, marketplace sellers and overseas “sourcing partners”. If they can commit the company to a supply chain, they are in scope for Kittel governance — not only payroll IR35.
2. Onboard the human, not only the supplier. For high-risk categories, run the same depth on introducers as on counterparties: identity, insolvency, prior directorships, disqualification, adverse media, and any known HMRC or tribunal history. In Eurolaser, reasonable steps that would have surfaced Darr’s prior MTIC involvement were part of the story HMRC and the tribunals cared about.
3. Define red-flag halt rules in writing. Examples: new counterparty above a value threshold without site or goods verification; identical lot sizes repeating through multiple “independent” suppliers; payment instructions that diverge from invoice entity; refusal to allow inspection; prices detached from public benchmarks. Halt means no PO, no payment, no input-tax booking until cleared by tax/finance — not “note and proceed”.
4. Separate commercial approval from VAT recovery assumption. ERP should not auto-recover input VAT on high-risk chains merely because a valid-looking VAT invoice exists. Park recovery, or require dual sign-off, until the chain file is complete.
5. Keep contemporaneous challenge notes. When a deal looks odd and you still proceed, write why the only reasonable explanation was not fraud. Tribunals reconstruct state of mind from the file you had then, not the narrative built after assessment.
6. Train procurement and treasury together. Fraud chains often need both a purchase story and a payment story. Treasury seeing circular bank flows or third-party payees is a control, not a courtesy copy.
7. Align SPA warranties and portfolio monitoring. For PE, add Kittel/open VAT fraud enquiry reps on buy-side VDD for wholesale, tech grey market, labour and logistics. Tell the audit committee residual risk in one slide: sectors, third-party buying models, open HMRC correspondence, introducer KYC quality.
What this is not
Attribution still turns on role, authority and engagement facts — not every consultant’s private knowledge destroys every claim, and back-to-back trading can be legitimate. It is still a reminder that VAT recovery depends on the integrity of the chain and the people who run it. “The director did not know” is incomplete if the company’s deal engine did.
CFO actions this month
- Inventory third parties who source or commit purchases in VAT-bearing goods or labour chains; flag self-employed introducers with book-of-business economics.
- Sample the last 12 months of high-risk categories for back-to-back concentration, repeat lots and thin files.
- Refresh introducer due diligence where onboarding was lighter than supplier onboarding.
- Confirm who can block a payment on a tax red flag without commercial override theatre.
- If HMRC has already written on supply-chain fraud risk, raise the diligence standard from that date — tribunals treat post-warning blindness harshly.
- Put one paragraph in the next audit-committee tax paper: Kittel residual risk, not only rate and partial-exemption maths.
Bottom line
Eurolaser [2026] UKUT 324 does not rewrite Kittel. It hardens the operational reading CFOs keep under-weighting: knowledge can sit in the consultant who builds the book, and still bind the company that takes the margin and the input tax. Director innocence, self-employed labels and “that’s normal in our market” are weak shields once the cumulative hallmarks and the deal-arranger’s sophistication are on the table.
Read the UT decision, keep VATF52100–VATF53420 and VATF73000 next to procurement policy, and treat introducer KYC as part of cash and tax control — not a sales ops nicety. If your growth model depends on a rainmaker who “knows everyone in the chain”, make sure finance knows who that rainmaker is.
General information for finance leaders, not advice on any specific supply chain, assessment or appeal. Take formal advice on live HMRC enquiries and tribunal strategy.