The conventional wisdom is that people struggle with money because they don’t understand it well enough. But when Cleo looked at the data, we found something more complicated.

In our survey of 10,000 adults aged 28–40 in the US and UK, 88% of respondents felt confident managing their finances. Yet about 23% reported both feeling confident and losing sleep over money.

That gap between knowledge and peace of mind is one of the most important problems we’re solving at Cleo. The financial services industry has spent years trying to educate people into better habits—and, often, implicitly blaming consumers for not understanding personal finance well enough when they can’t achieve their goals.

But for people who already understand their finances, education may not be the right lever. If financial literacy alone doesn’t bring peace of mind, what kind of support actually makes a difference? To answer that question, we asked people about their relationship with money and whether AI tools might have a role to play in improving it.

Where people want help from AI

You might expect a gradient, with people willing to trust AI for low-stakes tasks but drawing the line at anything consequential. Instead, the data suggests that people are asking a more binary question: “Am I okay with AI acting on my money, or not?”

Those who say yes don’t distinguish much by task type. Around half are comfortable with bill payment, overdraft assistance, savings transfers, and investing. Advisory functions like estimating disposable income or helping hit a savings goal score a bit higher, in the range of 55 to 58%.

Comfort levels span just 10 percentage points across task types, from investing (48%) to disposable income advice (58%).

We also asked what would help build users’ trust in AI financial assistants and found that people prioritize direct experience over external validation. Track record, regulatory oversight, and recommendations from people they know all mattered less than hands-on proof. The leading answer (22%) was the opportunity to start small and see results.

It’s not a knowledge problem

When we asked respondents about their biggest challenges managing money, knowledge gaps ranked at the bottom. Only 13% said they didn’t know how much they should be saving, the lowest of any challenge we asked about. The most common challenges fell into two different categories entirely.

Lack of knowledge ("not knowing how much to save") ranked lowest, at 13%.

The first is execution: putting knowledge into practice. Even if you’ve heard that you should spend less than you earn, or save a little toward retirement each month, acting on that advice is another matter.

Thirty-eight percent of respondents reported difficulty balancing today against tomorrow, while 37% referenced self-discipline. Another 18% cited not monitoring their spending, which follows the same pattern: Most people know they should track where their money goes, but struggle to do so consistently.

The second category is volatility: income and expenses that don’t arrive on a predictable schedule. The single most-cited challenge (43%) was unpredictable expenses like surprise bills and unplanned repairs. Another 21% cited unpredictable income, possibly from flex scheduling or gig work. And 27% pointed to supporting family or friends financially, another external pressure that’s erratic and hard to plan around.

None of these represents a lack of information or understanding. Many people know what they should do to achieve financial stability; the difficulty is in doing it week after week under uncertain conditions. You can read all the explainers in the world and still get knocked off course by a flat tire or a slow week as a rideshare driver.

23% of respondents reported both feeling confident about money and losing sleep over it.

This also helps explain why nearly a quarter of respondents felt confident about managing their money while also losing sleep over it. That combination makes more sense when you consider what people are up against. Confidence in your knowledge doesn’t help much when the problem is consistently executing your plans or adapting to circumstances you can’t predict.

This reframes what it means to help people with money. Financial literacy is a foundation of financial success, but education alone can’t solve the problem. AI financial assistants are a support layer that bridges the gap between intention and action. They can respond to new developments faster than someone checking their bank account every few days, and they’re built to help you follow through even when it’s hard.

How we’re building around this

This is the thinking behind Autopilot, Cleo’s most recent release. Autopilot leads with insight: showing you what you can afford, where your money is going, and what’s coming up next. For many users, getting insights and recommendations is the natural starting point before deciding whether to let Cleo do more.

But if the core problems are execution and volatility, guidance only gets you part of the way. Knowing you should move $50 to savings this week doesn’t help if you forget to do it, or if an unexpected medical bill throws off your plan before you get the chance.

This is why Autopilot also acts. Cleo can move money to savings, offer a cash advance to avoid overdrafts, adjust your budget when circumstances change, and help you follow through on goals you’ve already set. And she does it all with you watching, keeping you informed about what she’s doing and why.

Respondents told us they’re open to AI taking action, but they want visibility into what’s happening and clarity on the rationale. The goal is a system working alongside you that can act more quickly than you could on your own, while keeping you in the loop. People who understand their finances but are still stressed about them don’t need another budgeting tutorial. They do need an assistant who can help them follow through and can absorb the volatility they can’t control.

Most people are smarter about money than the financial industry gives them credit for. Cleo’s goal is to give them the tools to act on what they know and to insulate them against the shocks they can’t predict.

Data source: Cleo internal data as of January 2026.