Fraudsters are constantly evolving. From romance scams to deepfishing and ‘quishing’ – phishing using fake QR codes – criminal gangs are becoming increasingly sophisticated in how they target victims. Fraud remains the most common form of crime in the UK, costing society an estimated £14.4 billion annually and affecting more than four million victims each year.
To keep pace, banks are having to innovate just as quickly. Lloyds Banking Group, the UK’s largest digital bank, has long been at the forefront of fraud prevention, investing in new technologies and partnering with law enforcement to crack down on scams as they emerge. In 2025, the group prevented over £1bn in attempted fraud. Key to this are innovative features such as its upcoming Scam Check tool, which combines with other technology to flag and challenge suspicious transactions before they’re completed. Now the bank is taking a step further by pairing its expert colleagues with agentic artificial intelligence to strengthen its response to fraud.
The new tool, Fraud AI Agent, has been purpose-built to help Lloyds’ frontline support teams in dealing with fraud cases. When a debit card transaction is blocked, customers will still talk to a human – but that person is now assisted by an AI tool that brings together multiple response systems into a single streamlined semi-autonomous process. The AI agent helps fraud specialists assess factors including identity confidence and evaluates suspect transactions, while automating processes such as canceling and reissuing cards. The result? Faster, more effective and more consistent fraud response for customers at a moment when clarity and reassurance matter most.
Importantly, while others have embraced AI chatbots, Fraud AI Agent is designed with clear oversight and escalation points to ensure colleagues are able to audit, intervene, and make final decisions where needed. That ensures the system is safer, while freeing up the real people to provide a more personal customer experience.
Lloyds Banking Group has designed the architecture behind Fraud AI Agent to be ported and adapted across other customer experiences in the near future, particularly high-risk areas that share similar characteristics. It’s part of the group’s ambitious plan to roll out responsible “AI for All”.
To support this goal, Lloyds Banking Group has created an AI Centre of Excellence; launched its own AI Academy, intended to empower its 67,000 staff in building, using, and deploying AI in their roles; and has already trained 300 of its senior leaders in AI through an executive education programme with the University of Cambridge. More than 40,000 of Lloyds’ colleagues are already using Copilot tools in their everyday work, including delivering code written using Copilot Studio and Agent Builder. As a result of this and other initiatives, the Group has rolled out more than 50 generative AI use cases, ranging from customer service to risk assessment. All told, Lloyds estimates that generative AI is already delivering more than £50m a year in benefits to its business – a figure expected to double to £100m in 2026.
“AI is now integrated across our organization and is transforming how we operate and serve customers. Our ongoing investment keeps Lloyds Banking Group ahead in responsible tech adoption within financial services,” says Lloyds group chief operating officer Ron van Kemenade.
As more and more of us adopt AI in our everyday life, responsible use is going to be crucial to ensure the best outcomes. That’s why Lloyds is focused on putting its workers and customers first. Rather than replacing people with AI, it is equipping its colleagues with AI tools as part of a bold bet that they can innovate faster than the fraudsters – and deliver better experiences for customers with AI agents on their side.
