AI chatbots out-trust human scammers
Thursday 30th July, 2026
A study found AI chatbots outperformed humans at building trust in week‑long romance‑scam simulations, making victims likelier to install apps or follow requests.
This gives Wise a chance to comment on fraud risks in the UK and Europe and reassure customers about safe cross‑border payments and scam detection.
2 Sources:
Why have we flagged this?
This study shows LLM-driven scams are more effective at building trust and could increase sophisticated fraud targeting cross-border payments customers, so Wise can offer expert comment on fraud trends and protective measures.
Get reactive idea inspiration
Thinking...
These are starting points for inspiration, not finished ideas.
Data idea
‘Love’ transfers spike before first big loss
Wise shows that in suspected romance scams, customers often make a series of small, cross-border test transfers in the 10–14 days before a large loss.
Data to investigate: Analyse your internal scam case data: for confirmed or refunded romance-scam cases, chart the value and frequency of transfers in the 30 days prior to the report to quantify the small-to-large pattern.
Why: A single, concrete pattern from real payments gives reporters a fresh, defensible stat that makes the AI-enabled grooming phase tangible.
Expert comment
AI romance scams are about to industrialise fraud
Wise’s head of financial crime warns that LLMs will scale “pig butchering” globally by automating trust-building, and sets out three concrete defences banks must deploy now.
Expert comment angle: Offer a punchy op-ed and broadcast quotes from Wise’s Head of Financial Crime to BBC, FT, Sky News: AI shifts fraud from artisanal to industrial, so verification, payment friction at first-time payees, and cross-border data sharing must be mandated.
Why: A strong, contrarian line from a credible payments brand gives journalists a clear takeaway and policy prescriptions tied directly to the study.