Most people assume the youngest business owners are the ones racing ahead with AI. They grew up online. They picked up new apps the way earlier generations learned to ride a bike. So it feels natural to expect that a woman in her twenties running a small shop or studio would already have AI woven through her week. A new report from the JPMorgan Chase Institute complicates that picture, and the Gen Z women AI adoption gap it describes turns out to be the widest of any age group.
What the data actually shows
The report tracked de-identified Chase Business Banking accounts from 2019 to 2025, looking at which owners started paying for AI tools. Adoption climbed for everyone. In 2019, roughly 2 percent of men-owned businesses and 1.7 percent of women-owned businesses used AI. By 2025 those figures had reached 19.7 percent and 17.2 percent. The overall gender gap grew from 0.3 percentage points to 2.5.
The generational detail is the part worth sitting with. Among Gen Z owners, 20 percent of men-owned businesses had adopted AI against just 13.9 percent of women-owned ones. That is a 6.1 point gap, far larger than the 0.9 points among Gen X owners or the 0.5 among Baby Boomers. Millennial women trailed too, at 19.8 percent against 23.2 percent for men. Most striking of all, Gen Z women owners were less likely to have adopted AI than either men or women in Gen X. The digital-native generation is not leading here. On this measure, its women are furthest back.
Being online is not the same as being set up to build
It is tempting to read a number like that as a confidence problem. The evidence points somewhere more structural. Fluency with consumer apps, the thing Gen Z is genuinely good at, is not the same as having the time, the space to learn, and the operational control to fold AI into how a business actually runs. Those are different muscles.
Prowess, a UK organization that supports women in business, has described the same pattern among female entrepreneurs: the barrier is rarely interest and more often bandwidth, unpaid caring loads, and thinner professional networks to learn from. The JPMorgan authors note that women owners tend to report sharper concerns about data security, privacy and trustworthiness, concerns that land hardest on newer and younger businesses with less of a safety net behind them. A hesitation rooted in a real worry about client data is not a deficit. It is a reasonable response to a tool that still asks for a lot of trust.
A more useful question than confidence
The more useful question is not whether young women feel ready. It is what conditions let anyone integrate AI into daily work at all. Framed that way, the fixes look practical rather than personal. The report’s own recommendations point in that direction: targeted training aimed specifically at younger female entrepreneurs, and community-based mentorship where owners can talk through the messy parts of adoption with people facing the same constraints. That is the honest version of this story. The gap is not evidence that women are behind on ability. It reflects who gets handed time and permission to experiment, and who does not. This connects to a wider pattern in how women-owned businesses adopt AI fast but capture less of the payoff, and to the growing set of AI skills programs built specifically for women.
One small thing that counts
None of this needs to start with a grand plan. For an owner who has been meaning to try and keeps running out of week, the smallest useful move is to pick one task that repeats, one invoice reminder or one draft of a customer reply, and hand just that to a tool for a few days. Keep the client data out of it at first if that lowers the worry. Starting at all still puts a business ahead of where it was, and it makes the next step smaller.
The number that should stay in mind is the one about Gen Z women trailing Gen X women. It suggests the gap is not closing on its own as a younger, more online generation comes up. If anything it is being rebuilt. The question that leaves is uncomfortable but worth asking out loud: who is quietly being left to work out AI alone, and what would it take to change that before the gap sets?
