Romance and Investment Scams: The Long-Con Economy and What Banks Can Still Do About It

Romance and investment scams — increasingly merged into the 'pig-butchering' typology — have become the defining fraud problem of the decade. Losses per victim are often measured in six figures, and the human toll is frequently catastrophic. These scams are also evolving faster than most traditional controls can keep up with, incorporating deepfake video, AI-generated personas, and scripted emotional manipulation tuned to specific demographics.

For banks, the frustrating reality is that most of these payments are technically authorised by the customer. Traditional fraud controls are poorly calibrated to stop a victim willingly sending their own money to a criminal. Yet the costs — reputational, regulatory, and moral — fall on the institution. The result is that scam losses increasingly look like authorised, legitimate customer behaviour right up to the point where the funds cross into the criminal economy, which is precisely the space where detection is hardest.

The question for 2026 is no longer whether banks should intervene in long-con typologies, but how aggressively and with what supporting analytics. Institutions that wait for victims to complain have already lost the battle; the only viable model is proactive intervention informed by behavioural signals.

Regulatory, Enforcement, and Market Context

The UK, Australia, and Singapore have all moved furthest in creating reimbursement frameworks and intervention expectations. The Payment Systems Regulator's APP fraud reimbursement regime has reshaped the economics of customer warnings in the UK, and APRA and AUSTRAC have reinforced this in Australia. Reimbursement frameworks are reshaping the economics: when banks bear the downstream cost of victim losses, the business case for proactive intervention becomes obvious almost overnight.

FATF and the Egmont Group have both flagged the convergence between romance scams, scam centres, and human trafficking, emphasising the need for cross-border cooperation. OFAC has increasingly sanctioned networks involved in investment-fraud infrastructures operating from specific jurisdictions. Regulators are also increasingly interested in how effectively banks coordinate with telecom providers and social-media platforms, because the origin of many scams sits well outside the financial sector.

Regulators now expect banks to quantify their scam-detection effectiveness, not merely describe their controls. That shift has meaningful operational implications. The supervisory message is that scam response is no longer an internal fraud matter — it is an ecosystem problem in which banks have both exposure and leverage.

What the Data Is Showing

Chainalysis has repeatedly documented billions of dollars flowing from victims into crypto wallets linked to investment scams, with a significant concentration in Southeast Asian networks. Sumsub reports rising identity reuse and persona recycling across scam platforms. Chainalysis has also documented how scam proceeds increasingly route through a small number of off-ramp clusters, which opens a genuine opportunity for targeted choke-point interventions if the industry chooses to coordinate.

Reuters coverage of UK PSR data shows that while total APP fraud volumes have plateaued, losses per victim continue to climb — indicating that the remaining scams are longer, more targeted, and more sophisticated. Sumsub data on persona recycling is particularly revealing: the same synthetic identities show up across multiple platforms and jurisdictions, often for months before being detected — a pattern that data sharing could materially shorten.

Implications for Financial Institutions

Banks should treat romance and investment fraud as a first-class scenario in transaction monitoring, with specific signals around new payee creation, crypto exchange exposure, and emotional engagement indicators. Static rules do not capture this behaviour well. Effective intervention requires the ability to explain to a customer, in real time, why a specific payment concerns the institution — and that requires both analytics and trained conversational capability on the frontline.

Customer intervention must be designed carefully. Friction that is too soft fails to change behaviour; friction that is too aggressive drives victims around the bank entirely. The best programmes combine targeted warnings with trained branch and contact-centre staff. Where regulators allow data-sharing between banks, those partnerships repeatedly show material uplift in detection and victim-protection outcomes, and they are rapidly becoming a baseline supervisory expectation.

Long-term, industry-wide data sharing is the single biggest lever. The institutions that participate in scam-data cooperatives materially outperform those that do not. Investment in staff training — particularly in empathy-led conversation techniques — has been shown to be as important as the underlying detection technology in preventing completed scam payments.

Conclusion

Romance and investment fraud are not solvable by any one institution. But banks that invest in behavioural analytics, thoughtful customer intervention, and shared intelligence can break the grip these typologies have on vulnerable customers — and, increasingly, satisfy the supervisory bar being set. The message to the industry is direct: romance and investment scams reward collective action, and institutions that participate early benefit disproportionately.

Suggested Next Steps

• Build pig-butchering-specific scenarios into transaction monitoring and case prioritisation.

• Design layered customer warnings and coach frontline staff on intervention conversations.

• Participate in industry scam-data consortia where permitted by law.

• Track scam effectiveness metrics at the board level alongside traditional fraud losses.

Sources: UK Payment Systems Regulator, APRA, AUSTRAC, FATF, Egmont Group, OFAC designations, Chainalysis, Sumsub, Reuters.

TrustSphere helps financial institutions design and deploy intelligent fraud and financial crime detection solutions. Visit www.trustsphere.ai