A customer checks their bank account after paying for a flight. A few days later, a credit card offer appears, highlighting airport lounge access and travel rewards. Another customer regularly transfers money into savings and receives a message about an investment product. Someone browsing insurance plans gets a reminder explaining how much cover might cost each month.
These interactions are not necessarily new. Banks and financial companies have used customer information to target offers for years. What is changing is how quickly artificial intelligence can analyse that information, identify a potential need and help produce the message that follows.
Across banking, insurance, payments and wealth management, AI is becoming part of the machinery behind financial marketing. It is being used to identify prospective customers, personalise communication, recommend products and respond to enquiries. In some cases, it is also helping determine whether a customer should receive a sales message at all.
The shift comes as consumers increasingly turn to AI for financial information themselves.
McKinsey’s 2026 Global Banking Annual Review, based on a survey of approximately 46,000 retail banking customers across 13 markets, found that 23% were using generative AI for banking-related tasks. In India, the figure reached 53% among roughly 2,500 respondents. Among those using AI for banking, 74% accessed tools such as ChatGPT, Claude or Gemini.
For financial marketers, this presents two challenges. Customers are becoming more comfortable asking AI about money, while financial institutions are using the same technology to influence what customers discover and consider buying.
From mass campaigns to messages based on customer behaviour
Consider a bank promoting a personal loan.
A conventional campaign might target salaried customers within a particular age or income bracket. An AI-assisted system can potentially go further, examining existing relationships, previous responses, digital activity and the context of a customer’s interaction before recommending a suitable communication.
The difference is not simply that the message can be written faster. It is that the institution can test more variations, identify patterns across customer groups and adjust its approach as new information becomes available.
HDFC Bank has been experimenting with this approach in India.
At the ET MarTech+ Summit in October 2026, Deepak Oram of HDFC Bank discussed an analysis involving approximately one billion marketing messages. The bank examined how different wording, products and personalisation techniques related to campaign performance.
Some findings challenged familiar marketing assumptions. Messages using a customer’s name performed 24% below the median, while those referring to an account number performed 7% above it. Messages emphasising cheaper loan rates or lower EMIs performed 25% below the median, whereas messages communicating a product’s monthly cost performed 44% above it.
These were findings from one bank’s analysis, not universal rules for financial advertising. They nevertheless illustrate why institutions are looking beyond conventional customer segmentation.
Oram also highlighted the difficulty of turning analytical findings into changes across a large organisation.
“AI will throw data at you at a crazy pace, but if you don’t solve for organisational process, that insight will never get action.”
The challenge extends to how marketing teams work with product, technology and customer-service departments.
Kotak Mahindra Bank has described an approach that brings predictive modelling, customer behaviour, campaign measurement and AI-generated communication into a more connected marketing process.
Subhra Bardhan, SVP, Marketing and Head of Consumer Insights and Measurement at Kotak Mahindra Bank, explained the broader objective: “It’s not about communication. It’s more than communication. It’s connecting every dot, from curated experiences and products to touch points.”
That distinction matters because the next interaction may not be an advertisement.
ICICI Bank’s 2025-26 annual reporting describes in-house generative AI tools that analyse customer calls, emails and chats, connect those insights to customer profiles and identify contextual service or product opportunities. The bank also uses AI-generated email responses, subject to human validation.
A customer asking about an existing loan might therefore receive assistance relevant to that enquiry, rather than another generic promotion. Whether such recommendations are appropriate depends on the quality of the underlying data and the controls governing their use.
What happens when customers ask AI instead of searching for a bank?
AI is changing financial marketing from the customer’s side as well.
Someone comparing credit cards may no longer visit several websites to examine rewards and annual fees. Instead, they might ask an AI assistant to identify options suited to their monthly spending. A first-time investor could request a comparison of investment platforms, while a family looking for insurance might ask which policy features deserve attention.
This creates a different kind of competition for financial brands.
A product can be prominently advertised and still be absent from the AI-generated answer that a potential customer reads. Conversely, a clear explanation of fees, eligibility, benefits and limitations may become increasingly important when AI tools summarise financial options.
McKinsey’s research found that 62% of consumers using AI for banking would prefer AI-powered services and advice from their primary bank. Yet approximately half indicated they might consider a third-party financial application if their bank could not provide capable AI services.
Existing customer relationships therefore remain valuable, but they may not guarantee that a bank controls the beginning of the buying journey.
Deloitte’s 2026 survey of nearly 2,600 US banking customers adds another dimension. Among respondents who used generative AI, 58% had used it to research banking products. However, only 46% of all respondents trusted the accuracy of banking recommendations generated by AI tools, compared with 79% who trusted information on banks’ own websites.
The figures suggest that AI is gaining a role in discovery faster than it is earning complete trust.
Financial companies are responding by building AI into their own customer channels.
In 2026, Visa introduced an AI Financial Assistant designed for financial institutions to integrate into banking applications. It combines cardholder behaviour, relevant information and institution-held data to provide personalised financial guidance and surface offers. US pilots began in August.
Discovery Bank in South Africa has also introduced AI-assisted recommendations through its application and WhatsApp. According to a customer case study published by Microsoft, engagement with the bank’s personalised next-best actions doubled following the introduction of Discovery AI. The figure is company-reported and does not independently establish how much of the increase was caused by AI.
Stuart Emslie, Head of Actuarial and Data Science at Discovery Bank, described the ambition as giving customers “a private banker in their pocket”.
In India, similar developments are appearing across financial categories.
ICICI Prudential Life Insurance’s 2026 Partner Stack 2.0 includes an AI-based product recommender that suggests insurance options according to customer segment and life stage. Advisers can also generate personalised quotations to share with prospective policyholders.
In payments, PhonePe introduced SmartPages in June 2026, allowing merchants to describe their businesses in vernacular languages and use AI to generate payment-page content, fields and layouts.
These examples cover different stages of the customer journey. Some help people discover products. Others support sales conversations, produce marketing material or simplify the steps between interest and payment.
They also show why AI in finance marketing cannot be reduced to automated advertising. Increasingly, the marketing experience is embedded in the product or service itself.
Customers want personalisation, but how much data is too much?
The appeal of personalised financial communication is straightforward. A customer who is saving for a house may find a relevant home-loan explanation more useful than an unrelated credit-card promotion.
But financial data can reveal far more than shopping preferences. Transactions may indicate where someone travels, what services they use, how frequently they borrow and whether they are experiencing financial pressure.
That makes the use of AI in financial marketing more sensitive than personalisation in many other sectors.
Mastercard and FT Longitude surveyed 8,000 consumers and 300 senior executives across 11 markets in 2026. Seventy per cent of consumers said they were prepared to switch financial providers for more personalised financial guidance, while 42% reported that they had already begun switching for that reason.
The same study found that 70% were willing to share financial data securely. That willingness, however, was linked to understanding how their information would be used.
In practice, there is an important difference between using a customer’s stated preferences to explain a product and using sensitive behavioural information to encourage additional borrowing.
The distinction becomes especially significant when AI systems recommend financial products automatically or generate persuasive messages at scale.
Deloitte’s US banking survey found that 72% of respondents were concerned about sharing details of their financial situation with generative AI. Another 64% worried about hidden bias, while 83% said they would feel anxious about an AI agent taking financial action without their approval.
These concerns are not limited to data collection. They also involve the accuracy of recommendations and whether customers understand that AI is influencing an interaction.
India’s financial institutions face an additional challenge: different customers want different levels of digital assistance.
EY India’s 2026 study of 2,030 banking customers across urban, semi-urban, semi-rural and rural areas found that 55% wanted better digital support across banking applications, websites and chatbots.
Among the study’s customer groups, 49% of Rising Professionals and 51% of entrepreneurs prioritised smart digital features, including AI-powered financial advice and budgeting support. Yet half of the pre-retirement Golden Transitioners preferred speaking to a person.
For marketers, these differences complicate the idea that every customer journey should become fully automated.
A younger customer may welcome an AI assistant comparing savings products. Another customer may prefer a human adviser before making a long-term investment or insurance decision.
Regulators are also paying attention.
The Reserve Bank of India’s 2025 FREE-AI framework emphasises fairness, accountability, transparency and human oversight in financial AI. It identifies risks involving bias, privacy, inaccurate outputs and explainability, while making clear that institutions remain responsible for systems they deploy.
The UK Financial Conduct Authority has similarly examined the implications of increasingly autonomous AI in finance. Its 2026 consumer research, involving more than 5,000 UK retail-finance consumers, found that 20% would be likely to use AI capable of acting within pre-set goals.
FCA Executive Director Sheldon Mills said: “Artificial intelligence will transform financial services by 2030. It creates significant opportunities for consumers, firms and the wider economy.”
The regulator has also warned that poorly controlled adoption could amplify consumer harm, including misleading information and unfair outcomes.
For financial marketers, the practical implications are substantial. An AI-generated message about an insurance policy must accurately describe the cover. A loan promotion cannot invent an interest rate. An investment recommendation needs to reflect the relevant risks and limitations, not merely the customer’s likelihood of responding.
Human review, approved product information and clear accountability therefore remain important even as content production becomes faster.
The wider change is already visible. AI is helping financial institutions move from broad campaigns towards more contextual conversations, while customers increasingly use AI to evaluate the products being promoted to them.
But the success of that shift will not be measured only through click rates, leads or conversions.
In finance, a recommendation can affect someone’s savings, debt or financial security. As AI becomes more involved in deciding what customers see next, banks and financial companies will also have to demonstrate why those messages are appropriate, how customer information is being used and who is responsible when a recommendation is wrong.
Disclaimer: This article may include information derived from interviews, press releases, public statements, research, company communications and other publicly available or third-party sources. Such material may be summarised, paraphrased or contextualised for journalistic and editorial purposes. All rights in third-party content remain with their respective owners.