Accountability at Digital Speed

Framing remarks originally delivered at the Federal Reserve Bank of New York’s fifth annual “An Economy That Works for All — Financial Inclusion” event on 1/14/26

For decades, adult financial education has been positioned as a primary lever for financial inclusion. The intent is good, but the evidence has been mixed. Meta-analyses of financial education programs find occasional improvements in some contexts, but effects are small and inconsistent. At its worst, financial education reinforces a narrative that poverty is primarily a matter of personal responsibility, overlooking how systems, product design, and markets shape household outcomes. In a digital system that can monetize confusion at scale, the only serious answer is embedding clarity, guardrails, and accountability into products and infrastructure. 

Traditional Financial Education Under-Delivers

Financial education for adults is static, generic, and deficit-oriented — implicitly treating knowledge gaps as the main obstacle, rather than acknowledging structural conditions that shape household financial outcomes: wage volatility and uneven access to safe products.  

Even if we assumed stronger results, the old model can’t keep up with today’s market. Products are increasingly complex. The pace of innovation moves at record speed, so those already complex products can change rapidly. Both those facts deepen information asymmetries between businesses and the consumer. And much of the information consumers have access to is marketing optimized to shape choices at scale.

Finally, digital finance adds layers that financial education was never built to govern — privacy, security, explainability, reliability, and contestability when something goes wrong. When these risks materialize, consumers are often left without the visibility or leverage to understand what happened — let alone to contest and correct it.

The goal isn’t information for the sake of information; information is a means to achieve the actual goal of better outcomes. And to achieve that, education must be a part of infrastructure. To be clear, I’m not arguing for faster, better, cheaper digital financial education. I’m arguing for something more foundational: education as part of financial infrastructure. 

At its best, financial education turns good intentions into confident choices. Those same principals have to be embedded into the products, policies, and infrastructure people rely on every day. In practice, that will look very different from traditional education.

Modern digital financial services can operationalize what we’ve long known — outcomes improve when support is:

• Personalized, because they meet people where they are; 

• Contextual, grounded in both constraints and goals; 

• Dynamic, delivered when trade-offs are weighed; and

• Accountable, measured against financial well-being, not engagement metrics.

In plain terms, financial education for adults isn’t obsolete: it’s being upgraded, relocated, and embedded. Concretely, that means shifting from broad curricula to in-the-moment guidance on real trade-offs, from passive learning to active infrastructure that supports follow-through, and from a default story of “individual responsibility” to system-level accountability for whether the tools people use actually improve outcomes.

The Value-Add Test

Here’s my benchmark for meaningful innovation: A fintech product only adds value if it improves durable outcomes — savings, stability, long-term financial security. For people living with tight margins, the leading indicators are concrete:  

• Does it reduce volatility, timing gaps, or cash flow mismatches? 

• Does it expand agency, offering transparency and real choice?

• Does it advance goals with dignity, no hoops/bells, not shame?

If fintech can’t show progress on whether its use will advance outcomes, it’s not innovation — it’s just extraction, faster.

Embedding Education into Financial Tools: Four Practical Domains

Let me make this concrete across four domains — spending, credit, banking, and savings. 

Spending: Spending tools are increasingly automated. Consumers are interacting with tools that categorize and track spending, manage subscriptions, send bill reminders, and increasingly initiate actions — from payments to transfers — inside apps. Yet traditional financial education is no match for dark patterns, like features that intentionally hide opt-outs, or other design principles where the user experience is purposefully engineered to increase spending. It can’t outpace design choices engineered to increase spend, subscriptions, or churn.

Instead, embedded decision support should help people align spending with their values and trade-offs. Providers already use machine learning and emerging agentic tools for fraud detection, identity screening, transaction monitoring, and spotting AI-generated deepfakes. Looking ahead, it could also mean that consumers leverage collective buying power by using agents that aggregate demand for discounts or negotiate on a consumer’s behalf.  

Credit: Credit is increasingly digital and instantaneous. Consumers now encounter online applications, pre-qualification flows, and comparison tools — alongside underwriting that draws on cash-flow data from bank accounts, rent, or utilities. Traditional education isn’t built for a world where offers can be personalized, priced, and repriced dynamically — using data and model logic consumers can’t see — and decisions that happen in seconds. It can’t realistically prepare people to evaluate opaque trade-offs across multiple consent screens and “instant” approvals, especially when the costs show up later.

Traditional financial education’s response would be to have consumers read more disclosures. Embedded decision support, however, should make credit timely, understandable, and contestable by design — and support follow-through toward safer options.

Banking: Consumers are increasingly interacting with banking through digital channels — online onboarding and remote verification, account servicing and alerts, chat-based support, and in-app authentication. On the horizon are portable digital identity and reusable verification/KYC, alongside emerging “know your agent” expectations that clarify which tools — or agents — are authorized to act.

People can’t be educated out of deepfakes, identity theft, or rapidly changing threat models — especially when attackers can scale automation faster than individual learning. And traditional education isn’t designed for an environment where identity and authorization are continuous: multiple apps, conflicting permissions, and unclear responsibility when something goes wrong. As technology evolves the financial education “lesson” can’t be to memorize scam tips.

An education upgrade is ideal for people who are overbanked but underserved by leveraging stronger onboarding to provide more responsive services.  It must be to make protection actionable, goal-matched, and reinforced over time — so consumers can manage identity and authorization across platforms without becoming security experts.

Savings: Savings is where the convergence of innovation and embedded education could be most powerful, especially as new rails enable more programmable payments, including stablecoin-based arrangements. Consumers are increasingly using tools that make saving more automatic, like round-ups and auto-transfers.

Traditional education isn’t built for “if/then” financial systems where multiple rules interact, especially when liquidity is tight and timing is everything. As saving becomes more automated and more programmable, the “lesson” to embed is not to save $400, one and done. It’s to make tradeoffs visible, adjustable, and reinforced over time — so automation supports real constraints and real goals. We should be able to expect that when people can and will save, their choices become the path of least resistance — not a daily test of willpower. 

A concrete application on the horizon: As 530A accounts come into view, they’re a near-term test of whether modern digital infrastructure can help a public policy investment actually build household security by translating these principles into a mainstream savings vehicle. In this four-year pilot, roughly 14 million kids will be eligible for a $1,000 seed deposit.

If these accounts are positioned as universal basic accounts, they could support kids across life stages. That means designing them not as a one-time enrollment event, but as an ongoing relationship: Automatic enrollment; easy, low-friction ongoing contributions; protections so balances don’t trigger asset limits and unintentionally disrupt eligibility for public benefits. In sum, 530A accounts are a once-in-a-generation chance to modernize savings — pairing digital wallets and modern payment rails so more families can build security across life stages — not just a one-time enrollment event.

What It Takes To Make This Work At Scale

To ensure the benefits accrue to households, five conditions matter:

Business model alignment: Consumers who live paycheck-to-paycheck represent a $127B marketplace — one that deserves more fintechs than the handful that serve them today. The question is what business models will shape the next wave: When revenue depends on churn, overdrafts, or data extraction, financial education becomes theater. That’s why I worry about incentives to monetize the data of financially vulnerable consumers  in ways they can’t see or control: the “shadow costs” of extraction — privacy loss, manipulation, unfair pricing — have to be treated as real costs, not externalities. But when revenue is aligned with long-term success, innovation can compound outcomes and trust.

Legal and regulatory clarity that can keep pace: Digital systems are increasingly complex wherein multiple actors — engineers, users, and the technology itself — dynamically interact, each adapting, updating, and changing outputs in response to one another. In that environment, “human in the loop” is not sufficient if the system isn’t designed so that a person can genuinely oversee, interact with, or control its operation. A more scalable path is performance-based consumer protection: Require firms to implement systems that review the tool’s operations and performance, scrutinize the data the model relies on, monitor ongoing use, and audit outcomes for error, bias, equity, and inclusion.

Trust and implementation infrastructure: Uptake of digital tools can be slower when fraud risk is high and redress is weak. For families living with tight margins, reliability is not a “nice to have.” It’s the condition for participation at scale — clear liability, strong protections against fraud and error, and pathways to contest and correct. More broadly, digital inclusion, risk innovation, and ethical governance are now intertwined imperatives:  Trust has to be treated as an active design goal — measurable, governed, and improved. If we solve for business models and incentives that genuinely serve people living paycheck-to-paycheck, trust will follow.

Build stronger, shared evidence of what works: We need better ways to evaluate digital financial products using outcomes data (often controlled by private firms) so we can see what is helping households and what is harming them. That means encouraging data-sharing agreements between researchers and service providers, and public-private partnerships that allow for credible monitoring of use and outcomes.

Strengthen coordination as payment rails modernize: Where new payment rails or tokenization-related products create cross-cutting risks, the system benefits from an ongoing coordination mechanism across regulators — so risks are identified early and responses don’t lag the market. Without a standing forum, risk signals can surface in fragments and action can come late; tighter, routine coordination helps shorten that gap — especially for fast-moving arrangements like stablecoin-based payments and other tokenization-related activity.

Expect Different Inputs. Demand Better Results

We only have to look at other digital markets to see what happens when systems optimize for engagement rather than well-being. Financial services is too consequential to repeat that mistake. We must expect different inputs, and demand better results. 

Financial education doesn’t need to be discarded — but it must be upgraded and embedded into tools and systems that are dynamic, contextual, and accountable. The measure of success is not app downloads or access alone; it is financial security that endures for people navigating trade-offs every day.

In an economy that works for all, inclusion isn’t the marketing version — it’s the kind that reshapes business models and unlocks entirely new markets. The opportunity before us is not to explain volatility to the very people who experience it. It’s to reduce it. And prove we did.

The opinions shared in this article are the author’s own and do not reflect the views of any organization they are affiliated with.

Open Banker curates and shares policy perspectives in the evolving landscape of financial services for free.

If an idea matters, you’ll find it here. If you find an idea here, it matters. 

Interested in contributing to Open Banker? Send us an email at [email protected].

Leadership Can’t Be Automated

AI can help you move faster, but real leadership still requires human judgment.

The free resource 5 Traits AI Can’t Replace explains the traits leaders must protect in an AI-driven world and why BELAY Executive Assistants are built to support them.

Download BELAY’s Guide!