The Library Media — Science, Technology & Health · AI News & Intelligence Desk
The AI Bill Is Becoming a Utility Bill — and the Receipts Are More Complicated Than the Headline
By Gideon Black
Reuters framed the change in a striking way on November 13: AI has become a utility bill for Gen Z. In its “On the Money” report, Reuters described younger users stacking multiple AI subscriptions, sometimes spending hundreds of dollars a month, and said some would cut streaming services before giving up AI when household budgets tightened.
It is a compelling story. It is also a story in which the most memorable numbers come from individual behaviour, while the strongest evidence comes from a much broader category that does not isolate AI.
That distinction matters. The claim is not absurd. The receipts do show that subscription spending is growing, that Gen Z is leading that growth, and that reading-and-information subscriptions are expanding unusually quickly. But the receipts do not establish that AI alone is driving the change, that hundreds of dollars a month is becoming normal, or that younger consumers have collectively decided AI is more essential than entertainment.
The utility-bill framing is ahead of the evidence. Not necessarily wrong, but ahead.
Start with the anecdote, not the average
Reuters’ central profile was a 28-year-old video-ad studio operator spending $248 per month on Claude, ChatGPT, ElevenLabs and Grok, plus about $100 a year on Gemini.
That is a substantial subscription stack. It is also one person’s spending.
Reuters separately reported that power users were spending $200 or more each month across AI applications. The report also described users who said they would cut Netflix or Spotify before AI if their budgets came under pressure.
Those observations are worth reporting. They reveal how some people are incorporating AI services into their working lives and how they rank those services against other discretionary spending.
But a profile is not a population estimate. A power user is not a representative consumer simply because the example is vivid. And a person’s stated intention to cancel streaming before AI is not the same as observed behaviour when money actually gets tight.
The distinction is especially important when the headline invites readers to infer a generational shift from a handful of unusually committed users.
Reuters has identified a real kind of consumer: someone for whom several AI services have become sufficiently useful to justify recurring payments. It has not, through those profiles, established how common that consumer is.
Ryan Marshall, a partner at ELA Financial Group, offered Reuters a personal-user rule of thumb: spending above roughly $20 a month may be overspending for personal use. That is an attributed opinion, not a measured threshold, and it should not be mistaken for a financial benchmark derived from the bank data.
The question is not whether $248 is a lot of money. For many households, it plainly would be a meaningful monthly expense. The question is whether that spending describes a trend across Gen Z or a particularly visible edge of the market.
For that, the bank data deserves more attention.
What the bank actually counted
The Bank of America Institute’s September 2026 report, Consumers subscribe to more spending, is built from anonymized Bank of America payments data. Its figures offer a broader view of subscription behaviour than a collection of individual profiles.
According to the Institute, overall subscription spending rose 7.7% year over year in July 2026. That growth outpaced overall card spending growth by more than 1.5 percentage points, a pattern the Institute said had persisted for the past two years.
Gen Z posted the fastest subscription spending growth: nearly 14% year over year in July 2026.
Those are meaningful findings. They support the proposition that younger Bank of America cardholders are increasing their subscription spending faster than other groups, and that subscriptions are gaining ground relative to overall card spending.
But neither figure is an AI statistic.
The 14% measures Gen Z subscription spending across categories. It does not measure spending on ChatGPT, Claude, Gemini or any other AI product. It does not tell us how much of the increase came from AI subscriptions, or even how many Gen Z cardholders paid for an AI service.
The Institute’s more specific category is “reading and information.” That category includes AI-related offerings alongside traditional media such as magazines, news and audiobooks.
It is the smallest share of subscription spending, according to the Institute, but the fastest growing. In July 2026, spending in the category rose 61% year over year for Gen Z and 51% for younger Millennials.
Those are striking growth rates. They are also category-level rates, not AI-specific rates.
The category’s inclusion of AI makes it relevant to the question. It does not make the category synonymous with AI.
A reader who sees “61% growth” and concludes that Gen Z’s AI spending rose 61% has been given a claim the data does not support. The number describes a basket containing several kinds of reading and information services. The Institute does not break out AI alone.
This is not a technical footnote. It is the central limitation on using the report to validate Reuters’ framing.
The less fashionable subscriptions still dominate
There is another finding in the Bank of America Institute report that complicates the idea that AI is already becoming the defining subscription expense.
The Institute says entertainment and retail subscriptions drove most of the overall increase. Together, they account for roughly 43% of all subscription spending.
That is a useful corrective to a narrative built around AI’s rapid ascent.
The fastest-growing category is not necessarily the largest. A small category can post extraordinary percentage growth while remaining a relatively small part of total household expenditure.
And the category that includes AI-related services is, according to the Institute, the smallest share of subscription spending.
That does not make its growth unimportant. Early changes in a small category can matter, particularly if they continue. But it does mean the broader subscription economy remains substantially shaped by entertainment and retail.
The utility-bill metaphor risks obscuring that scale distinction.
A service can become important to a growing group of users without becoming a dominant household expense. It can become indispensable to a particular professional without becoming a generational necessity. And it can grow rapidly from a small base without yet representing a large share of total spending.
The bank’s figures support growth. They do not establish dominance.
The receipts have limitations of their own
Bank of America Institute’s data is stronger evidence of spending patterns than a single consumer profile, but it is not a complete census of American subscription behaviour.
The Institute’s figures describe Bank of America cardholders. The Institute also notes possible selection bias.
That matters because cardholder behaviour in one banking population cannot automatically be treated as the behaviour of all Gen Z consumers. The data provides a substantial view of transactions within the population observed, but the reader should not silently expand that population into an entire generation.
There is also a measurement problem inside the word “growth.”
The Institute’s figures capture spending growth, which can reflect both price increases and new subscriptions. An increase in dollars spent does not necessarily mean an equivalent increase in the number of services purchased, users subscribing or hours of use.
If an existing subscription becomes more expensive, spending can rise without a new service being adopted. If more consumers subscribe, spending can rise for a different reason. The reported figures combine those possibilities.
The distinction matters when trying to infer what is happening to consumer habits.
A rise in subscription spending is evidence that consumers are paying more for subscriptions in aggregate. It is not, by itself, evidence that they are adding more AI products, using them more frequently or treating them as essential.
The report gives us a sound basis for saying that subscriptions are taking a larger share of spending growth among the observed cardholders. It gives us a much weaker basis for saying why.
What survives scrutiny?
Several parts of the viral claim survive, if stated carefully.
First, subscription spending is growing faster than overall card spending in the Bank of America Institute’s data. The pattern has persisted for two years, according to the Institute.
Second, Gen Z has the fastest subscription spending growth among the groups discussed, at nearly 14% year over year in July 2026.
Third, the reading-and-information category that includes AI-related services is growing particularly quickly: 61% for Gen Z and 51% for younger Millennials, according to the Institute.
Fourth, Reuters’ reporting provides concrete examples of consumers who have assembled expensive AI subscription stacks and who say they would preserve those services ahead of streaming subscriptions.
Together, these facts justify taking the change seriously.
They suggest that AI services are entering the subscription budgets of some consumers, including some younger users, and that the broader category containing AI is expanding quickly.
But several stronger claims do not survive.
The data does not show that AI spending among Gen Z rose 61%. It does not show that 14% subscription growth was driven by AI. It does not establish that a $200-plus monthly AI stack is typical. It does not establish that most younger users would cancel Netflix or Spotify before AI. And it does not show that AI has displaced entertainment and retail as the main forces behind subscription spending growth.
The distinction between “some users are behaving this way” and “a generation is behaving this way” is not pedantry. It is the difference between a reported trend and a manufactured consensus.
Is AI becoming a utility?
The metaphor gets one important thing right.
For some users, AI is no longer an occasional novelty purchased out of curiosity. Reuters’ profiles show people paying for several services because those services have become part of how they work. When a recurring service becomes embedded in a person’s workflow, cancelling it can feel less like removing entertainment and more like removing a tool.
That is the utility-bill intuition: recurring expenditure tied to ongoing use, difficult to abandon when the service has become part of everyday activity.
But the metaphor also smuggles in a stronger claim.
A utility is not simply something a person uses regularly or values highly. Calling AI a utility suggests a degree of necessity and normalisation that the evidence here does not establish across Gen Z.
The Bank of America Institute’s figures show subscription growth, not universal dependence. Its fastest-growing category includes AI but does not isolate it. Reuters’ examples show intense individual commitment, not a generational average. And the spending data cannot tell us whether the increase comes from more subscriptions, higher prices or both.
My verdict: AI is becoming a recurring budget item for some consumers, and the category containing AI-related subscriptions is growing quickly. That is already worth watching.
But the evidence does not yet justify saying that AI has become a utility bill for Gen Z as a whole.
The more accurate headline would be less viral and more useful: younger consumers are increasing subscription spending, while a small but fast-growing information category—including AI—is gaining ground.
That is a real story. It does not need a larger claim attached to it.
The skeptic should neither dismiss the change because the evidence is imperfect nor inflate the evidence because the change is exciting. The receipts show a market moving. They do not yet show exactly where it is going.
Gideon Black
Chief Skeptic, The Library Media
From The Library — the estate of our family.
