NEW YORK — Artificial intelligence has reached nearly every corner of American small business, with 86 percent of surveyed firms now using it, according to a Goldman Sachs report published in October 2026 — but only a small share have woven it into the core of how they operate, leaving a wide gap between trying AI and truly depending on it.
Illustration: A small business owner using a laptop at a shop counter — an artist's impression, not a photograph of a survey respondent.
What Happened
Goldman Sachs published its October 2026 report, titled "The AI Economy Is a Small Business Economy," based on a survey of 2,473 alumni of the bank's 10,000 Small Businesses programme in the United States, according to coverage by ANI on October 24, 2026. The headline finding is striking: AI adoption among the surveyed firms rose to 86 percent in 2026, up from 56 percent in 2024 and 77 percent in 2025 — a 30-percentage-point jump in two years.
The detail underneath the headline is more cautious. Only 15 percent of all respondents said they had fully integrated AI into their core operations, ANI reported. Among firms already using AI, the fully integrated share was 18 percent — the two figures describe different groups, so both can be true at once: most firms have tried AI, but fewer than one in five users has made it central to the business. "The gap between adoption and integration is now the defining frontier for small businesses," the report stated, as quoted by ANI. "High adoption does not mean deep integration," it added.
The survey also found that 93 percent of AI users reported a positive impact from the technology, with 74 percent citing efficiency and time savings, according to NewKerala's October 24 coverage of the report. At the same time, 74 percent of AI users said they had no formal policy governing how employees use AI — a governance gap that grows more consequential as the tools spread from the front counter into finance, hiring and customer data.
Alongside the report, Goldman Sachs announced on October 22, 2026, a $500 million commitment to expand its 10,000 Small Businesses programme and launch a Small Business AI Institute, according to reporting on the announcement. The institute brings together Anthropic, Google, Meta and OpenAI to support AI education and adoption among American entrepreneurs, with training, research on real-world use, and work on expanding access to capital.
Illustration: A lively main street of small shops — an artist's impression, not a photograph of a particular town.
Background
Small businesses are the quiet majority of the American economy: they employ roughly half the private workforce and account for a large share of new jobs, even though each firm is small on its own. That is why a survey of programme alumni matters beyond its sample — if the firms that already receive training and support still struggle to integrate AI deeply, firms without that support are likely to struggle more. But it is also why the survey's limits matter: the respondents are graduates of one bank's entrepreneurship programme, not a random sample of every small business in the country, so the 86 percent figure should not be read as the adoption rate for all American small firms.
The shift in how firms use AI is as telling as the adoption rate. The share of AI-using businesses employing the technology for financial analysis and business planning more than doubled, from 18 percent in 2024 to 42 percent in 2026, ANI reported, while AI use for data analysis rose from 26 percent to 42 percent over the same period. Use for written marketing content, by contrast, fell from 74 percent to 65 percent. In other words, AI is moving from the marketing department into the back office — from writing social posts to helping decide what to stock, what to charge and where the money goes.
That move raises the stakes. A chatbot that drafts a mediocre advertisement costs a firm a little embarrassment. A system that shapes financial planning, pricing or credit decisions can cost real money if it is wrong, biased or fed bad data. The survey's finding that most users lack a formal AI policy therefore reads less like a paperwork complaint and more like a risk warning: the tools are being trusted with more consequential work before the rules for using them have caught up.
A Babson College survey accompanying the announcement, as reported by Brief.News, adds a practical detail: 77 percent of graduates wanted more training and resources, and 73 percent spent less than $100 a month on AI, with many relying on free tools. That suggests the barrier is not only willingness but also know-how and budget. Free tools are enough to experiment with; turning an experiment into a dependable daily process usually takes training, time and someone on staff who understands both the business and the tool.
Why It Matters
The adoption-versus-integration gap is the story that policymakers, lenders and technology companies should actually plan around. If 86 percent adoption were the whole story, the small-business AI revolution would already be over, and the only remaining question would be which vendor wins. The integration figures say something different: the revolution is still in its trial phase, and the winners will be decided by training, trust and practical support rather than by who shouts loudest about AI.
For lenders and investors, the gap cuts both ways. Firms that integrate AI into finance and operations may become more efficient and more creditworthy — but firms that adopt AI tools without policies, training or data discipline may take on new risks that do not show up in a traditional loan application. Goldman's own $500 million commitment, with its emphasis on education and access to capital, reads as a bet that the bottleneck is practical capability, not enthusiasm. If that bet is right, the next wave of small-business growth will come less from new models and more from boring, essential work: teaching owners what to automate, what to keep human, and how to check the machine's homework.
For workers, the shift from marketing uses toward financial and operational decisions matters too. When AI helps plan a rota, forecast demand or screen applicants, it touches people's hours, pay and prospects in ways a marketing draft never did. The absence of formal policies at three-quarters of AI-using firms means those decisions are often being shaped by informal habits — one employee's favourite tool, one manager's prompt — rather than by agreed rules. As AI use deepens, employees, customers and regulators will reasonably ask for clearer answers about how those decisions are made.
Illustration: Two bar charts, one tall and one short, symbolising the gap between AI adoption and full integration — an artist's impression, not survey graphics.
What Happens Next
The Small Business AI Institute's first test will be whether training translates into integration. Its programme, as described in reporting on the announcement, combines partner-led training, research on real-world small-business AI use, and work on access to capital — the three things the survey suggests firms actually lack. Watch for concrete measures in the coming months: how many firms complete the training, whether graduates report higher integration rates in the next survey round, and whether lenders treat AI capability as a factor in credit decisions.
Expect the policy gap to close from both directions. Some firms will write their first AI-use policies once a tool touches payroll, customer data or hiring — often after a near-miss rather than before one. Regulators and industry groups, meanwhile, are likely to publish practical guidance aimed at firms too small to hire a compliance team. The firms that move first on simple rules — what data may go into a public tool, who checks AI-generated figures, and how customers are told when AI is involved — will have an advantage when those expectations harden into norms.
The broader economy will also keep score. If integration rises, productivity gains that have so far been concentrated in large companies could spread to the small firms that employ so many people — a genuinely broad-based effect. If integration stalls, the 86 percent headline will look, in hindsight, like a wave that broke early: lots of experimentation, little transformation. Either way, the next survey will be more interesting than this one, because it will show whether the defining frontier moved — or stayed exactly where Goldman Sachs found it.
Sponsored: explore today's featured sponsored offers
FAQ
What did Goldman Sachs actually survey?
A survey of 2,473 alumni of its 10,000 Small Businesses programme in the United States, not a random sample of all American small businesses, according to ANI's October 24, 2026 coverage.
How many small businesses use AI?
Among surveyed firms, 86 percent in 2026, up from 56 percent in 2024 and 77 percent in 2025, according to the October 2026 report "The AI Economy Is a Small Business Economy."
Why are there two different "fully integrated" figures?
Because they use different denominators: 15 percent of all respondents said they had fully integrated AI into core operations, while 18 percent of firms already using AI reported full integration, as reported by ANI.
What are firms actually using AI for now?
Increasingly for financial analysis, business planning and data analysis — each used by 42 percent of AI-using firms in 2026, up sharply from 2024 — while written marketing content use fell from 74 percent to 65 percent.
What is the Small Business AI Institute?
A new initiative announced by Goldman Sachs on October 22, 2026, backed by a $500 million commitment, bringing together Anthropic, Google, Meta and OpenAI to support AI education, research and access to capital for small businesses, as reported in coverage of the announcement.
Sources
- Goldman Sachs, "The AI Economy Is a Small Business Economy," October 2026 report, as covered by ANI, October 24, 2026.
- NewKerala, "Goldman Sachs: US small business AI adoption hits 86 pc," October 24, 2026.
- Brief.News, "Goldman Sachs Launches AI Institute to Boost Small Business Innovation," October 22, 2026 (Babson College survey findings as reported there).
Comments
Post a Comment