Happy third birthday, ChatGPT: How AI changed banking forever

Three years ago, on 30^th November 2022, ChatGPT was released and the world was given an artificial intelligence (AI)-powered large language model (LLM) that can understand and respond to human language by processing vast amounts of multimedia data.

Within five days of its release in 2022, one million people had tried ChatGPT; within two months the figure had reached 100 million. It was the fastest-ever mass adoption of a consumer product.

In this instalment of Finextra’s Explainer series, we explore what ChatGPT has achieved in its first three years of life, and how AI is reshaping the financial services industry.

Three years of industrial revolution

Though the unique selling point of ChatGPT may not have sounded particularly revolutionary at the time, the implications have been seismic. Led by CEO Sam Altman, OpenAI – the AI research and deployment company behind ChatGPT – ushered in a new industrial revolution and technological paradigm – one that saw AI invested in or adopted by many of the world’s biggest companies and financial institutions (FIs).

The craze pushed Google to accelerate the development of its own AI tool, Google Gemini; it revitalised Microsoft’s market position; and it triggered an AI start-up and venture capital explosion. Indeed, entire industries have been raised on the promise of AI – from colossal chip manufacturers like Nvidia, to sprawling data centres across China and the US. Between 2024 and 2027, global spending on AI data centres is expected to exceed $1.4 trillion.

But the longevity of this ecosystem rests on whether such investments can deliver robust profits. Many big investors – including Peter Thiel and SoftBank – have been spooked by the threat of a bubble forming, and recently jettisoned their AI holdings. Nvidia – the lynchpin of the AI supply chain – remains resolute, its Q3 2025 earnings having beaten expectations. For now, the markets are reassured of the strength and demand for AI infrastructure.

AI’s promise-delivery gap

The true test for the AI industry is whether the AI-powered tools deployed by organisations can deliver tangible results – be it improved productivity, operational efficiency, streamlined compliance processes, or next-level customer experience.

While at the macro level, AI seems to be having little impact – with The Economist showing that the US has seen “almost zero economic productivity growth from AI” – the technology is demonstrating utility in discreet, specific use cases. The financial services industry, for its part, has been at the forefront of AI adoption since day dot and could offer a blueprint for other sectors looking to ensure their deployments deliver a meaningful return on investment.

Three banking applications

So, how are FIs using AI? Here are three ways in which institutions are already deploying this tool across their front and back-ends:

1. Virtual assistants

   
   

On the front end, AI is being used to serve customers and triage requests. Lloyds, for example, unveiled in November 2025 the UK’s first multi-feature AI-powered financial assistant, which will be available from early 2026. It is designed to help customers manage spending, savings, and investments, with plans to expand functionality across all Lloyds’ financial products.

2. Fraud monitoring

   
   

AI is also sharpening institutions’ fraud prevention capabilities. In September 2024, Mastercard expanded a first-of-its-kind AI technology that spots anomalies in transaction data – handing UK banks greater visibility over potentially fraudulent activity, so they can stop scams before they take place. By strengthening defences against financial crime, AI is also helping these FIs comply with regulations, especially anti-money laundering (AML), and know-your-customer (KYC) rules.

3. Personalisation

   
   

Thanks to AI’s ability to digest and rationalise vast amounts of consumer data – across transactions, marketing, financial history, and more – banks can use it to generate new cross-selling opportunities, issue timely offers of credit or personalise their communications. Essentially, AI is a means for banks to tailor their interface with customers and ensure they are supported at the right time, in the right ways – thus driving customer satisfaction and loyalty.

Implementations in all these areas are having a tangible impact. A recent PwC report showed that fully embracing AI can drive a 15-percentage-point improvement in banks’ efficiency ratios.

In the future, AI could be used to deliver hyper-personalised banking products with fully customisable features, next-generation investment advice, or automated teller services. With the right guardrails, the possibilities are almost endless.

A bifurcated outlook

At three years old, ChatGPT is too young to be judged for the consequences it brought about. What we can say is that the release of ChatGPT triggered a wave of AI-powered technology that swept the financial services industry – streamlining compliance processes and upscaling the customer experience, but at the same time handing criminals a sophisticated tool with which to defraud everyday consumers. Like most technologies, AI has no moral compass and can be used for good or bad.   

In decades to come, will we look back at ChatGPT’s birthday, 30^th November 2022, as the beginning of an industrial revolution that unlocked economic productivity, or will we mark it as a catalyst for job displacement, disinformation, cybersecurity issues, technological dependence, and socio-economic inequity? The answer to this question is being crystallised today – and hinges on how successfully providers can purge from their models opaque decisioning processes, biases, and hallucinations.