Nearly half of employer firms in the United States are now using artificial intelligence tools, and those companies report significantly brighter expectations for growth than their peers, according to a Federal Reserve Bank of New York study published October 8, 2026, on the bank's Liberty Street Economics blog. The research, which analyzed 5,248 employer firms from the 2025 Small Business Credit Survey, found that AI adopters scored higher on both employment and revenue outlook measures. The findings suggest a correlation between technology use and confidence, though the report emphasizes the link doesn't prove causation.
The survey revealed that 46% of employer firms said they or their staff were using AI tools, with an additional 15% planning to adopt within the next twelve months. Companies deploying AI registered a net employment expectation score of 33 points, compared to 15 points among non-users. Revenue outlook showed an even wider initial gap: AI users scored 48 points against 21 for companies not using the technology. When researchers adjusted for firm and owner traits and historical performance, the differences narrowed to roughly 10 points for employment outlook and 14 points for revenue expectations, but remained positive.
The authors caution that the study captures associations rather than proven cause-and-effect relationships, and measures anticipated outcomes instead of actual results. Among firms using AI, 77% reported no change in labor costs, and just 31% said they'd seen increased sales from AI adoption. The report notes that whether companies actually realize the growth they expect remains a question for future research. Among AI users, 51% had integrated the technology partially or fully into their business processes.
The gap between optimism and observed results points to a central challenge: anticipating AI's impact is easier than delivering it. Companies may feel confident about technology's potential while still working through the practical hurdles of deployment, training, and workflow integration. The relatively small share of firms reporting tangible sales gains suggests many are still in early stages of adoption, where upfront investment and learning curves precede measurable returns. The research positions AI as a tool to test against individual business metrics, not a guaranteed lever for expansion.
The report offers three concrete steps for small businesses: focus on one measurable task rather than broad adoption, establish a repeatable workflow before adding technology, and evaluate results based on actual sales or leads after a month rather than sentiment. The bottom line is that optimism about AI doesn't automatically translate into performance gains, and companies should treat the technology as an experiment requiring careful tracking and adjustment. For decision-makers, the real test isn't whether a tool feels promising but whether it shifts the numbers that matter to revenue and customer acquisition. The study's inability to separate enthusiasm from outcomes suggests businesses may be overestimating near-term returns while underinvesting in the process changes that make technology stick.