APP & Romance Fraud Detection
Understand the story behind a scam – and stop the losses before the final payment.
APP fraud and romance scams don’t start with a suspicious link. They start with trust, grooming and slowly escalating pressure. Our AI agents read the language in messages over time, spotting red flags of manipulation so people can get help sooner – avoiding life-changing losses – and helping teams make fast, fair, defensible decisions that reduce write-offs and investigation costs.
Individuals:
Organisations:
Why language-based fraud detection matters
Every year, thousands of people are manipulated into sending money to scammers they believe they can trust.
In the UK alone, police record around 8,800 romance fraud cases a year, with reported losses close to £95 million and average losses in the five-figure range.
Actual numbers are higher: shame and fear mean many victims never report.
For individuals, a single scam can wipe out savings; for banks and platforms, every missed red flag turns into refunds, write-offs and complaints. Catching coercive language early means fewer ruined finances – and fewer costly cases to resolve.
Traditional fraud tools focus on devices, locations and transaction patterns.
Those signals matter – but the real harm happens in the conversations that persuade someone to move their money.
The reassurance, the urgency, the emotional blackmail.
herEthical AI brings psychology into fraud detection.
We analyse patterns of coercion and grooming across chats, emails and call notes, so you can see how a scam unfolded – not just when the money left the account.
For individuals: “Am I being scammed?”
A quiet check before things escalate – and before big money leaves your account
If you’re worried about someone you’re talking to online or over the phone, you can paste or upload your conversations into our checker before you send money or share sensitive details.
Our agents look for patterns like:
-
manufactured urgency (“you have to act now or we lose everything”)
-
dependence and over-reassurance
-
guilt, shame or threats when you hesitate
-
sudden shifts from affection to pressure around money
You’ll see:
-
a clear summary of what’s happening in the messages
-
a timeline of red-flag events
-
plain-language explanations of why each instance looks risky
A few minutes of checking can be the difference between a worrying feeling and losing tens of thousands of pounds.
We don’t tell you what to do. We give you enough clarity to decide – so you can walk away from risky situations before they turn into real financial loss.
For banks, PSPs and investigators
Explainable evidence for hard calls
For financial institutions, APP fraud is both a human and regulatory challenge.
Banks in the UK are liable for reimbursing many APP fraud losses – up to £85k per case – within strict timelines, often just five days from notification. Every unclear case risks unnecessary reimbursement, complaints and manual review time. Our module helps you:
Our APP & romance fraud module helps you:
-
Triage cases faster and cut manual review time per claim
-
Build clear coercion timelines from message history to resolve disputes in hours, not days
-
Link language patterns to evidence for internal/external review without trawling through hundreds of screenshots
-
Evidence fair treatment under regulatory scrutiny and reduce the risk of costly challenges and fines
This is not another black-box score. Every flag comes with a human-readable explanation your teams can inspect, challenge and improve.
Even small gains add up: reducing unnecessary reimbursements and handling time across thousands of cases a year can translate into millions in protected revenue and avoided operational costs.
Scammers are increasingly using AI to sound convincing; we use AI to spot the manipulation early, so you lose fewer customers – and fewer pounds – to those scams.
01
Ingest evidence
Upload chat logs, emails or call notes – or ingest them automatically via API from your CRM, case management or dispute systems
02
Add witness statements
Upload any narrative descriptions provided by the fraud target or witnesses.
03
Detect patterns of grooming and coercion
Our agents scan messages over time to find key behaviours: emotional bonding, isolation, rehearsal of payments, pressure, threats, even sextortion.
Each event is labelled with its type and severity, not just flagged as “suspicious”.
04
Review a clear, explainable timeline
Investigators see a visual timeline of events, linked to exact excerpts in the underlying evidence. They can add notes, adjust labels, and export a complete report.
Start small, scale safely
Individual & small-team access
A secure web tool where people or support workers can:
-
paste or upload conversations
-
see a red-flag timeline and risk summary
-
download a simple report to share with banks, police or counsellors
Ideal for charities, victim services and advice lines who need to support clients but don’t have specialist fraud analysts.
Organisation licences
For banks, PSPs, insurers and law enforcement:
-
Workspace: browser-based console for fraud teams to review timelines, annotate cases, and bundle exportable reports.
-
API: integrate our detection engine into your fraud stack to pre-score cases, enrich transaction monitoring, or add coercion markers into your case management system.
-
Analytics: aggregate statistics across cases to pinpoint where scams are emerging, what tactics are being used, and where staff training or customer education will have most impact.
