Family Business Fraud: Three Siblings, One Deception-Indicated
£38,000 had been taken from the family building company their father had spent 25 years building. Three siblings — all directors — had financial access. All three were tested in a single day. Two came back clear. One came back deception-indicated. This case study includes the actual P300 EEG waveform data so you can see exactly what a deception response looks like — and what it doesn't.
Important Privacy & Limitations Notice
Anonymisation: All names, company identifiers, and identifying details have been fully anonymised.
EEG data: The waveform graphs below are constructed from anonymised session data consistent with the findings described. Amplitude values and timing parameters reflect the actual profile recorded in this investigation. Raw data in the written report uses the same format.
Purpose: For educational and investigative planning purposes. Not legal advice.
Case Background
"Hartley Building Supplies" (anonymised) was a family-run trade supplier in the East Midlands, founded by the siblings' father and passed to his three children on his death in 2021. "James" (eldest, financial oversight), "Rachel" (operations and procurement), and "Paul" (sales and customer accounts) each held a directorship and shared access to the business banking facilities.
In early 2026, the company's external accountant identified a pattern of irregular cash withdrawals and unrecognised supplier payments during the preparation of the annual accounts. The discrepancies were traced back approximately two years and totalled approximately £38,000 — a combination of cash withdrawals from the business account and payments to a supplier account that the accountant was unable to match to any purchase record or invoice in the company's files.
None of the three siblings initially had a satisfactory explanation. James pointed to Rachel's involvement in procurement. Rachel pointed to Paul's management of cash sales. Paul pointed to James's sole management of the online banking credentials. The company's solicitor advised that without an objective means of identifying the responsible party, the family faced the prospect of either extended legal action between siblings or writing off the loss without resolution. P300 EEG was instructed.
The P300 Response: What the Data Shows
Before presenting the case results, this section shows you what P300 EEG data actually looks like — and why the difference between a deception-indicated and a clear result is visible even to a non-specialist reading the raw waveform.
When a stimulus appears on screen, the brain produces a characteristic sequence of electrical responses. In a standard Concealed Information Test, the probe stimuli are details only the guilty person would recognise. If their brain recognises the probe, it produces a large positive voltage deflection peaking around 300 milliseconds — the P300 component. If they have no concealed knowledge of the probe content, no such peak appears.
Event-Related Potential (ERP) — Pz Electrode
Average waveform response to probe stimuli vs control stimuli · Time zero = stimulus onset · Positive values plotted upward (convention: inverted) · Shaded region = P300 analysis window (250–450ms)
Investigation Design
All three siblings were tested sequentially across a single day, each isolated from the others. The probe design was built around three specific categories of knowledge relating to the unrecognised supplier payments and the cash withdrawals — details that only the person who had made or directed them would hold in neurological memory.
Probe Set 1 — Knowledge of the unrecognised supplier account
Stimuli built around the specific supplier account — its name, its bank details, its connection to the person who controlled it. Only someone who had set up or used the account to route company funds would hold these specific details in the form tested. The probe included both the account's actual details (unknown to two of the three siblings) and plausible distractor accounts.
Probe Set 2 — Cash withdrawal mechanics and timing
Stimuli built around the specific pattern of the cash withdrawals — the amounts, the intervals, the method used to withdraw without triggering the company's informal oversight protocols. Only the person who had designed this pattern would hold it in neurological memory in the specific form the probe stimuli tested.
Probe Set 3 — Accumulation: scope and duration
Stimuli built around the approximate total accumulated over the two-year period — a figure known only to the accountant and the investigation team, not disclosed to any of the three siblings before or during the sessions. Recognition of this figure in the specific neurological form tested is consistent with having tracked it as an active participant.
Rachel — Session 1, morning
Three probe sets. 85-minute session. Rachel was visibly anxious at the briefing — concerned, she said, not about the result but about the process of testing her brothers. Her EEG baseline was stable throughout.
Paul — Session 2, midday
Three probe sets. 85-minute session. Paul attended with a notebook of his own financial records for the relevant period, which he offered unprompted at the briefing. He was told the specific probe design would not use his self-produced records. He agreed to proceed.
James — Session 3, afternoon
Three probe sets plus an extended Probe Set 1 (additional stimuli around the supplier account) based on a specific detail identified during the morning sessions. 100-minute session. James was calm throughout — similar to other cases in our dataset where the subject holds specific concealed knowledge and has made peace with the likelihood of a positive result.
Results delivered — company solicitor — same evening
All three results delivered jointly to the company solicitor and the two non-tested-first siblings in a single end-of-day debrief. Written reports for all three delivered the following morning.
Results
Rachel
Produced no meaningful P300 recognition responses across all three probe sets. Waveform profile consistent with no concealed knowledge of the supplier account, the cash withdrawal pattern, or the accumulated total. Cleared on the day of testing.
Paul
Produced no meaningful P300 recognition responses across all three probe sets. Waveform profile consistent with no concealed knowledge of any of the tested categories. Cleared on the day of testing.
James
Produced statistically significant P300 recognition responses across all three probe sets — probability scores of 96%, 92%, and 94% respectively. The strongest responses were on Probe Set 1 (supplier account) and Probe Set 3 (accumulated total). Waveform data shown above. Deception indicated.
What the waveform comparison shows — in plain language
The EEG graphs above are representative of what was recorded in James's session versus what was recorded in Rachel's and Paul's sessions. The difference is not subtle. When James saw the probe stimuli containing specific details of the supplier account and the cash withdrawal pattern, his brain produced a large positive electrical response at approximately 318 milliseconds — the P300 component. When Rachel and Paul saw the same stimuli, their brains produced waveforms indistinguishable from their response to control stimuli.
The P300 fires because the brain has encountered something it already knows. James's brain knew those details. Rachel's and Paul's did not. That difference is visible in the data before any probability calculation is applied — it is present in the raw waveform.
Key Investigation Findings
- James returned deception-indicated on all three probe sets, with probability scores of 92–96%. His strongest P300 response was on Probe Set 1 (supplier account), consistent with holding the specific details of the account's ownership and setup.
- Rachel and Paul both returned clear across all three probe sets. Their waveforms showed no statistically significant P300 response to any of the probe stimuli — matching the profile of subjects with no concealed knowledge of the fraud mechanics.
- The P300 waveform data for James showed a peak amplitude of 14.2 μV at 318ms on the probe condition — compared to 2.1 μV on the control condition and a probe-condition peak of 2.1 μV for the cleared siblings. This difference is visible in the raw data without further analysis.
- James admitted the diversion after being informed of his result, before leaving the building. His account confirmed that the supplier account was connected to a third-party individual he had a personal financial relationship with — details consistent with the Probe Set 1 recognition responses.
- The investigation resolved in a single working day a dispute that had already consumed weeks of solicitor time and had brought three siblings to the edge of formal legal action against each other.
Outcome
James admitted the fraud the same afternoon his result was delivered. He confirmed the supplier account had been set up by a third party at his direction, and that the cash withdrawals had been used to service a personal debt he had been carrying since shortly after their father's death. The total he admitted to was £38,000 — consistent with the accountant's analysis.
The company solicitor drew up a recovery agreement between the three siblings providing for James to repay the £38,000 from his director's drawing account and future dividend entitlement, in lieu of the civil proceedings that would otherwise have followed. James accepted the terms. The agreement avoided litigation, kept the business operational, and preserved — in as much as it could be preserved — the sibling relationship that the fraud had damaged.
Rachel and Paul each received written P300 EEG reports documenting their clear results. James received his deception-indicated report as part of the formal recovery documentation. All three reports were retained by the company solicitor as part of the matter file.
What This Case Demonstrates
Family business fraud has specific dynamics that conventional investigation cannot navigate
When the suspects are siblings and co-directors, conventional investigative approaches — HR processes, formal interviews, forensic accountancy — produce findings that each sibling can attribute to the other's management failures. The account discrepancies in this case were real and documented; who had caused them was not. P300 EEG answered the second question without requiring any of the three siblings to produce evidence against another — the neurological data spoke independently of any of them.
The waveform data makes the finding visible rather than just asserted
Most P300 EEG case studies describe a probability score and a conclusion. This case study includes the actual waveform comparison because the visual difference between James's response and his siblings' responses is the most direct illustration of why the technology works. The P300 peak at 318ms is not a model prediction or a statistical inference — it is an electrical measurement of his brain recognising specific details of the fraud it had stored. The cleared siblings produced no such measurement because they had nothing to recognise.
Resolution without litigation preserves what litigation destroys
The recovery agreement was possible because the P300 result gave the family a factual foundation that removed the dispute about who was responsible. Without it, the likely trajectory was civil litigation between siblings — which would have consumed the equivalent of the stolen amount in legal costs, destroyed the working relationships within the business, and potentially forced a sale of the company their father had built. The investigation cost a fraction of those consequences and resolved the matter in a single day.
Investigating Fraud in a Family Business?
P300 EEG identifies the responsible party — and clears those who aren't — in a single day. Initial consultation is free. Written report within 24 hours including raw EEG waveform data.