Curinos joined Bloomberg Intelligence for its second AI-focused webinar, this time turning to a question now front of mind for every deposit franchise: What happens to bank funding when artificial intelligence makes it dramatically easier for consumers to find, compare and move money between deposit accounts? Brian Buckingham, Senior Vice President of Consumer Deposits at Curinos, walked through the forces already reshaping deposit behavior, a three-phase framework for how AI is likely to land, and the practical steps banks can take now to protect — and sharpen — their value proposition. The throughline is that most of what matters is still within a bank’s control, provided leadership engages with it deliberately rather than waiting for the picture to come into perfect focus. Here are the key takeaways from the discussion.
1. Deposit inertia has protected bank funding for decades — and the friction behind it is starting to erode.
Deposits have long been one of the most valuable and stable parts of a bank’s franchise, and Curinos attributes that stability to three forms of friction that together create what we call deposit inertia: the attention required for a consumer to even be aware of better options; the effort of actually switching; and the materiality of the benefit, which for many simply hasn’t been worth the hassle. That friction has translated into stable, low-cost deposits and, in turn, low-cost funding for the asset side of the balance sheet. While much of the friction is still there, it’s being chipped away by a wider competitive set of fintechs and non-banks, easier digital onboarding, and technology shifts that loosen the historic link between the payment and the deposit. One signal of the shift: of consumers actively switching institutions, those with at least four checking relationships rose from 7% in 2019 to 50% in 2024 — evidence of fracturing wallets and a growing willingness to spread money around.
2. AI’s impact on deposits will arrive in three phases — productivity, capability and autonomy — and the industry is still early.
AI’s impact can’t be reduced to a single event; it’s better understood as a progression:
- Productivity — doing things that were going to get done, only faster. Summarizing a document, building an analysis, having Gen AI quickly scan and summarize the best posted deposit rates in the market.
- Capability — doing things that realistically couldn’t have been done before. For example, setting up an AI agent to continuously monitor the market within defined parameters — such as alerting when a six-month CD crosses a target rate.
- Autonomy — taking action on a consumer’s behalf. For example, fully agentic money movement: a target-balance “sweep” that holds a set amount in an account, then automatically finds the best rate, opens the account, moves the money and sweeps it back as needed, with the deposit holder largely out of the loop.
Most activity today sits in productivity and early capability. Autonomy remains further off and genuinely unresolved. That’s because today’s models reason probabilistically rather than like a human, which raises real questions about the guardrails required before consumers hand over decisions about their money (and regulators allow it).
3. The most exposed customers are the ones for whom rate already matters most.
Higher-balance and rate-sensitive segments — affluent and high-net-worth households, digitally engaged customers, large corporates and large uninsured balances — are where highest-rate-paid behavior already exists, and where AI tooling is most likely to accelerate. The opportunity becomes clearer at the margin: Curinos estimates that just under $2 trillion in savings balances are currently paid less than 100 basis points, while market acquisition rates sit well north of 200 bps. For the mass market customer with $5,000 to $7,000 in balances, the near-term benefit of AI is modest; but the adoption curve deserves respect. According to one report, it took roughly two years for 45% of working-age adults to adopt some form of generative AI, a level of engagement that took digital banking around 15 years to reach.
4. AI is becoming the gatekeeper to your brand — and optimizing for the model is the new SEO.
Consumers who use AI to make decisions everywhere else in their lives are also using it to evaluate financial institutions. That means the primary consumer of your online information is increasingly an LLM, not a person — and LLMs don’t read just your website. They scan Reddit threads, social posts and third-party reviews alongside what you say about yourself, then reduce it to roughly three recommendations, often with a single winner. This brings into focus two priorities. First, the aperture needs to be widened. Banks should actively manage the full body of online information about their institution, not just owned channels. Second, FIs need to be clear, concise and even a little reductive. They need to make unmistakable who they are, who they serve, how their products line up and what they cost, with pricing prominent. Nuance and complex narrative may be suitable for other conversations; here, machine-readable clarity wins the recommendation.
5. Primacy needs a sharper, more deliberate definition.
Primacy starts with self-knowledge. A bank committed to a relationship and cross-sell model must make that value proposition genuinely cohesive — rewarding the behavior it wants and eliminating the disconnect of treating well-known deposit customers like a stranger when they ask for a loan. A bank that’s comfortable holding the transactional relationship while a customer parks savings elsewhere needs to be just as deliberate: define exactly what you’re defending — the direct-deposit and payments anchor — and make sure the functional capabilities can sufficiently monetize beyond the core relationship. A mismatch between the stated value proposition and the lived experience can be fatal. While none of this changes overnight, the move toward clarifying the target customer and aligning product, pricing and message should start now.
6. Reduced search-and-switching friction can change deposit economics even before autonomy arrives — and new intermediaries are the ones to watch.
A simple reduction in friction can move the needle because it lowers awareness and effort barriers even while the materiality question remains. The development to watch is the potential entry of aggregator or “clearinghouse” players — in effect, Mint on steroids — that give a consumer a single view of balances across every institution and facilitate money movement between accounts they already own, surfacing the best available rates alongside. Larger banks are experimenting here as well. Reports suggest at least one money-center bank is beta-testing an agentic product that moves customer money into higher-yielding options automatically. Tellingly, such “easy button” propositions are designed to be high-yielding in value rather than necessarily highest in rate — which only underscores this point: a clear value proposition, well told, is what earns the right to manage a customer’s money.
7. The biggest opportunity is within the portfolio — break the silos and turn data into coordinated decisions.
The most important thing a bank can do now is to commit to a deep understanding of what its deposit customers actually care about — both rate and non-rate factors — and, critically, where rate matters, because it doesn’t matter evenly across regions, segments, balance tiers or products. Pay sharp where rate drives behavior; redeploy into non-rate value where it doesn’t. Then make sure the bank gets credit for it. A well-priced product no one hears about is a tree falling in an empty forest — it raises your marginal cost of funds without driving acquisition. Underpinning all of this is coordination. Local optimization among pricing, marketing and product destroys enterprise value. The path forward is a connected decision layer — the heart of Curinos’ approach to decision intelligence — that brings the right data to the right people at the right moment, so banks don’t just set a price but activate the right customers with the right offer.
