Fraudsters stole £1.2 billion through payment fraud in the United Kingdom in 2025, with artificial intelligence increasingly helping criminals to manipulate victims, impersonate trusted people and execute scams at scale.
The latest figures from UK Finance show that authorised push payment fraud — scams in which victims are tricked into sending money to criminals — rose sharply last year. Losses from APP fraud increased by 19 percent to £576.4 million, the highest level since 2021. Reported cases also rose by 7 percent to 248,070.
The development is not just a British banking story. It is a warning for Nigeria, where electronic payments have grown rapidly, fraud losses remain significant and criminals are already skilled at exploiting trust, social media, bank transfers and weak digital literacy.
AI Is Making Fraud More Convincing
For years, many scams were easy to detect: badly written emails, crude SMS messages, fake bank alerts and poorly designed investment websites.
Artificial intelligence is changing that. Fraudsters can now generate polished messages, imitate the tone of banks and fintech companies, clone voices, create fake images, automate conversations and produce convincing investment pitches in seconds.
This lowers the cost of fraud. It also allows criminals to target many more people while making each message appear personal.
In the UK, two-thirds of APP fraud cases originated online, showing how much modern fraud now begins on digital platforms before ending in bank transfers.
Nigeria’s Fraud Problem Is Already Large
Nigeria does not need to wait for AI-powered fraud to become a problem. The country already has a large and evolving electronic fraud challenge.
According to the Nigeria Inter-Bank Settlement System, actual digital payment fraud losses stood at about ₦17.67 billion in 2023. They then surged to ₦52.26 billion in 2024, partly because of a single large fraud incident of ₦31.1 billion involving one entity. In 2025, losses fell by 51 percent to ₦25.85 billion.
The decline is encouraging, but the figures still show that electronic fraud is a major cost to Nigeria’s banking and payments system.
FITC data also show how fraud is shifting across channels. Nigerian banks lost ₦3.3 billion to fraud in the first quarter of 2025, up from ₦1.39 billion in the previous quarter. Mobile fraud losses rose sharply, while computer and web fraud also increased.
This suggests that even when the number of fraud incidents declines, successful attacks can become more damaging.
Why AI Could Be Dangerous in Nigeria
Nigeria’s fraud ecosystem is already heavily based on social engineering. Criminals do not always need to hack a bank. They often persuade victims to give away passwords, one-time passcodes, card details or transfer money voluntarily.
AI could make this much worse.
A fraudster could clone the voice of a parent, spouse, employer or religious leader and request an urgent transfer. Fake customer-service agents could use AI-generated scripts to sound more professional. Criminals could create fake videos of celebrities or business leaders promoting investment schemes. WhatsApp and Telegram fraud groups could automate thousands of personalised messages.
Investment fraud is a particularly serious risk. In the UK, investment fraud losses rose by 40 percent to £221.5 million. Nigeria already has a history of Ponzi schemes, fake crypto platforms, forex trading scams and high-yield investment programmes. AI will make fake testimonials, fake trading dashboards and fake founder profiles easier to produce.
Romance fraud could also become more sophisticated. AI-generated photos, cloned voices and automated conversations can help criminals maintain false relationships for months, making emotional manipulation more effective.
The Next Phase of Nigerian Fraud
If Nigerian fraudsters adopt AI at scale, the next wave of fraud is likely to move in four directions.
First, impersonation scams will become more believable. Victims may receive calls or voice notes that appear to come from relatives, bank staff, company executives or public officials.
Second, investment scams will become more professional. Fraudulent schemes will have better websites, better videos, better pitch decks and more convincing social media campaigns.
Third, business email compromise may become more damaging. AI can help criminals imitate the writing style of company executives and send payment instructions that look authentic.
Fourth, customer-service fraud may expand. Fake bank, telco and fintech support accounts can use AI to respond quickly and convincingly to customers seeking help online.
Banks Cannot Fight This Alone
The UK debate has focused partly on reimbursement rules, because British banks are now required to refund many APP fraud victims unless the customer was grossly negligent.
Nigeria does not yet have an equivalent consumer-protection regime at that scale. This means prevention is even more important.
Banks and fintechs need stronger transaction monitoring, faster fraud-response systems and better identity checks. Telcos need to help stop SIM-swap and identity-related fraud. Social media platforms need to remove fake investment adverts and impersonation accounts more quickly. Regulators need to treat AI-enabled fraud as a financial-stability and consumer-protection issue.
The lesson from the UK is clear: fraud is no longer just a banking problem. It is a digital ecosystem problem.
For Nigeria, the risk is that AI could turn familiar scams into more convincing, faster and more profitable criminal operations. The country’s digital payments revolution has created enormous convenience. It has also created a larger battlefield for fraudsters.
