trending
Fraud in digital finance is no longer best understood as a series of isolated scams carried out by opportunistic criminals. It is increasingly operating as an industrialised system: fast, scalable and technologically sophisticated.
AI fraud is becoming industrialised
That is the central warning from Vyntra’s latest fraud trends report, which argues that the banking sector is confronting a new phase of criminal activity shaped by artificial intelligence, hyper-personalisation and real-time monetisation.
The headline figures are sobering. Vyntra estimates that global scam losses reached $442bn over the past 12 months, while 70 per cent of adults worldwide encountered at least one scam attempt and almost a quarter lost money.
Those numbers point not simply to a rising fraud problem, but to a structural threat to trust in digital payments and banking.
From manual deception to scaled AI-enabled fraud
What has changed is the speed and precision with which fraudsters can now operate.
Generative AI has dramatically reduced the time required to create convincing phishing messages, fake identities and impersonation campaigns.
According to the report, a process that once took more than 16 hours can now be completed in under five minutes.
That shift allows criminals to launch highly personalised attacks at industrial scale, targeting thousands of victims at once with messages tailored to appear credible and urgent.
The result is a shrinking intervention window for banks and payments providers.
Nearly two thirds of scams now succeed within a single day of first contact, leaving far less time for manual review or reactive fraud controls. In practice, that makes older approaches to fraud management look increasingly inadequate.
APP scams and social engineering remain at the centre
Among the most pressing threats are Authorised Push Payment scams, in which victims are manipulated into sending funds themselves.
These attacks are particularly difficult to prevent because the transaction often appears legitimate at the point of initiation. Vyntra also highlights the growing sophistication of phishing-enabled account takeover, executive impersonation, romance fraud, recruitment scams, QR code abuse and invoice manipulation.
Across these typologies, the pattern is consistent: criminals are combining AI-generated emails, voice cloning, deepfake video and spoofed identities to build trust quickly and move stolen funds before institutions can intervene.
Fraud is therefore becoming more than a payments issue; it is a broader challenge for compliance, operational resilience and customer protection.
Fraud prevention is becoming a collective intelligence problem
The report’s broader argument is that fraud prevention can no longer be treated as a siloed control function.
As instant payments accelerate the movement of money, financial institutions need real-time behavioural analytics, integrated transaction intelligence and structured information-sharing across the sector. Collaborative detection, rather than institution-by-institution response, is becoming essential.
That is especially important because the consequences extend beyond financial loss. Large-scale scam operations are increasingly linked to organised crime and, in some cases, human trafficking networks.
For banks, fintechs and payment providers, the implication is clear: fraud is no longer a peripheral operational risk. It is a systemic challenge that demands faster technology, deeper co-operation and a far more proactive defence model.
Comments