Parents pay more for AI homework tools when other teens use them

What the study tested

Source: University of Cologne via BlueSky Education, 15 September 2026.

A study led by Professor Christopher Roth of the University of Cologne has examined how parents value AI subscriptions for their children, and what happens to that valuation when the perceived number of teenage users changes. The research was published in the Proceedings of the National Academy of Sciences (PNAS) and surveyed more than 2,000 parents in the US, Canada and the UK. The design was straightforward. Parents were shown scenarios in which 20, 40, 60 or 80 percent of teenagers were said to use AI. They were then asked about their willingness to pay for an AI subscription for their own child. The results showed a clear pattern: as the assumed share of teenage users rose, so did the amount parents said they would pay.

The numbers are specific. For every 10 percentage point increase in assumed usage, willingness to pay rose by an average of 1.38 dollars. Between the 20 percent and 80 percent assumed usage scenarios, willingness to pay rose by more than 60 percent. That is a substantial shift driven purely by a change in the perceived social norm, not by any change in the product itself. The study therefore isolates peer perception as a distinct driver of parental spending. For UK households, the finding matters because it describes a mechanism that can push families towards yet another monthly subscription, independent of whether the tool is judged to be effective.

Fear of falling behind, not belief in the tool

The study found that the decisive factor for parents was that their child should not fall behind other teenagers. This was not primarily a matter of parents believing that AI tools improve learning. The driver was comparative: the wish to avoid a relative disadvantage. That distinction matters for how the market for consumer AI subscriptions is likely to behave. If demand is powered by a fear of falling behind, it can remain strong even when parents hold mixed views about the technology itself. The study captures this ambivalence directly. Parents were not uniformly enthusiastic about AI for children, yet their willingness to pay still tracked the perceived usage of other teenagers.

Professor Roth said: "It highlights that social forces operate with enough strength to sustain adoption of a technology that parents themselves view with ambivalence." The finding also helps explain why information about risks such as long-term learning losses did not change willingness to pay. Parents could absorb a warning about potential harms and still report the same willingness to pay for a subscription. The social comparison appears to sit apart from the risk assessment. For a UK household weighing a monthly AI subscription for a child, the study suggests the decision may be shaped more by what other families are believed to be doing than by a settled view on the tool's educational value.

Warnings shift attitudes but not wallets

The study did find that risk information changed attitudes in one respect. After risk information, support for school-wide rules or bans on AI rose from 44 percent to 57 percent. That is a meaningful shift in stated policy preference. But the same information left individual willingness to pay virtually unchanged. Parents became more supportive of collective restrictions while remaining just as ready to pay for a subscription for their own child. This combination is notable because it suggests that public and private preferences can move in different directions at the same time.

Professor Roth said: "Whilst the information about risks did change their attitudes towards AI as support for school-wide rules or bans increased from 44 to 57%, the willingness to pay individually for an AI subscription remained virtually unchanged." The pattern has a practical implication. Warnings aimed at individual purchasing decisions may not reduce take-up, even when they shift views about what schools should do. For UK families, the study indicates that a decision to buy an AI subscription is unlikely to be reversed by risk messaging alone. The social pressure to keep up appears to be the stronger force at the point of payment, while concerns about risk are expressed at the level of school policy.

What it means for UK household budgets

Consumer AI subscriptions are typically priced as monthly recurring payments. A family paying for one or more per child adds a standing cost alongside broadband and mobile. The study's finding that willingness to pay rises with perceived teenage usage points to a budget line that can grow through peer pressure rather than through a deliberate assessment of value. For UK households, that makes the subscription a recurring commitment rather than a one-off purchase. Several practical points follow. Check whether the school already provides a licensed tool, since paying separately may duplicate something already available. Use free tiers before paying, to test whether the tool is actually used. Set the subscription to renew manually, so the payment does not continue by default.

Consumer law provides some protection. The Consumer Contracts Regulations 2013 give a 14-day cancellation right on most online purchases including subscriptions. The Digital Markets, Competition and Consumers Act 2024 introduces rules on subscription reminders and easy cancellation. These measures matter more when a subscription is bought under social pressure, because the initial decision may not reflect a settled preference. A 14-day window allows a household to test whether the tool is used before the first renewal. Reminder and easy-exit rules reduce the risk of a subscription continuing unnoticed. Taken together, the study and the consumer rules suggest that the main budget risk is not the first payment but the renewal that follows a decision made to keep up with other families.

Related coverage on Kael Tripton: UK School Uniform Costs and Financial Help, UK Childcare Costs and Government Funding Explained, Mazda Launches Subscription Car Insurance With Wrisk, Checkmycar Cancel Subscription: UK Car Insurance Guide, Aviva Launches Home Insurance App on ChatGPT.

Frequently asked questions

How much did willingness to pay rise between the 20 percent and 80 percent usage scenarios?

Willingness to pay rose by more than 60 percent between the 20 percent and 80 percent assumed usage scenarios.

What was the increase for every 10 percentage point rise in assumed usage?

For every 10 percentage point increase in assumed usage, willingness to pay rose by an average of 1.38 dollars.

Who led the study and where was it published?

The study was led by Professor Christopher Roth of the University of Cologne and published in the Proceedings of the National Academy of Sciences (PNAS).

How many parents were surveyed and in which countries?

More than 2,000 parents were surveyed in the US, Canada and the UK.

Did risk information reduce willingness to pay?

No. Information about risks such as long-term learning losses did not change willingness to pay.