AI-generated article. This article was researched and drafted using AI tools and published automatically, and its featured image was generated by AI. Facts are drawn from the sources cited in the text.
October arrived with a label on it. National Women in AI Month runs through the month, and for a woman running a small business that lands somewhere between encouraging and irrelevant, depending on the week she is having.
What National Women in AI Month is actually for
National Day Calendar created the observance with the Cadence Giving Foundation in 2024, backed by a $20 million commitment and launched alongside the Fem.AI Summit. Its stated purpose is to showcase women as role models, promote their success in AI and tech, and encourage women into AI careers. The shortfall it points at is well documented. The World Economic Forum put women at roughly 26% of data and AI roles, against about 47% of the US labor force.
The month opened with a story worth noticing
On September 28, AMD announced it would acquire World Labs in an all-stock deal valued at $8.2 billion, as reported by CNBC and Bloomberg. World Labs was founded by Fei-Fei Li, the Stanford researcher behind ImageNet and co-founder of the university’s Institute for Human-Centered AI, to build world models, systems that construct and reason about three-dimensional environments. Li joins AMD as executive vice president and chief scientist, reporting to chief executive Lisa Su. Women’s Agenda pointed out what makes the deal unusual at this scale: a woman founder selling to a woman chief executive.
That is a real milestone. It is also about as far from most readers’ Tuesday as a story can get.
Visibility is the cheapest intervention on the list
Role models matter, and they are the least expensive thing anyone can offer. The research on what actually separates women from AI adaptation points at costs a month of profiles does not touch.
A working paper by Katelyn Cranney of Stanford, Solene Delecourt of Berkeley Haas and Rembrand Koning of Harvard Business School pooled 318,924 respondents across more than 100 countries and found men using generative AI at 47.8% against 39.3% for women. The gap narrowed after 2023, then stopped narrowing, holding near 16% in relative terms since early 2025. The authors sort the causes in two: frictions that wider diffusion will clear on its own, and institutional barriers that will not. The second group holds occupational segregation, unequal access to training and institutional support, sharper privacy and risk concerns, and the competence penalty women pay for visible AI use. Their conclusion about the remedy is blunt. Those barriers need direct organizational intervention, not patience.
The JPMorganChase Institute found the same shape in the population this site writes for. In analysis published in May 2026 from Chase business banking records through 2025, male-owned small businesses adopted AI at 19.7% against 17.2% for female-owned firms. That headline gap is narrow. The generational detail is not. Among Generation Z owners it widens to 6.1 percentage points, 20.0% against 13.9%, and the Institute names capital constraints, greater concern about data security and trustworthiness, and thinner support networks among younger women owners. The wider range in small business adoption statistics is worth reading alongside them.
The harder gap is depth, not sign-up
Adoption counts have stopped being the interesting question. Research published this year by the Cherie Blair Foundation for Women, with Intuit and the World Bank, surveyed more than 3,000 women business owners across low and middle income countries. Reported use jumped from 38% in 2024 to 82% in 2025, and nearly seven in ten reported saving time.
Then the detail that matters. Only about a third of frequent users applied AI to operations, and 35% to bookkeeping and finance. Marketing dominated. The report’s own framing is that the question is no longer whether women entrepreneurs adopt these tools, but whether they are equipped to use them deeply in the functions that decide whether a business lasts. It names time pressure, gaps in applied skills and the absence of structured implementation support, which is the same stall described in fast adoption that never quite embeds.
A smaller bar for the next few weeks
None of this becomes a program anyone can start on a Thursday afternoon. It does point somewhere.
Most reported use in these studies sits in marketing, which is the lowest-risk place to put it and also where gains are easiest to overstate. The more useful move is to take one operational task, the invoice chase, the stock count, the weekly numbers, and spend twenty minutes seeing whether a tool already paid for can take a first pass at it. One task. Not a stack.
That is a deliberately low bar, and it is still progress. The Cherie Blair figures make the point plainly: women entrepreneurs have adopted these tools at scale and are still using them in one corner of the business, which says less about any individual than about how quickly the learning curve compressed. When we talk to these systems the answers arrive complete, and they still have to be read, checked and absorbed. That work is more tiring than it sounds.
A month of role models gives women in AI something they have been short of, which is visibility. Whether it reaches a woman with four employees and no spare hour depends on what gets built after the profiles are published: who pays for the training, who protects the time, and who decides she is allowed to change how the work gets done. That part is not on the calendar.
