Big companies are scaling AI agents. Small businesses are stuck at 22%.

Are AI agents only for big companies? The newest numbers say the gap is growing. The good news is that it does not have to.

McKinsey's 2026 State of AI survey found that among large companies (over $1 billion in revenue), the share scaling AI agents jumped from 27% to 40% in one year. For smaller organizations, it stayed flat at 22%.

At the same time, small businesses are not sitting still. A New York Fed analysis published on October 8 found that 46% of small employer firms already use AI tools, and another 15% plan to start within a year. AI users were also more optimistic about growth: their net revenue expectations were 48 points, compared with 21 for non-users. The authors are careful to say this shows a connection, not proof that AI caused it.

So small businesses are trying AI. What most are not doing yet is turning it into agents that run real workflows.

Why the gap?

In my work with business owners, the problem is rarely the tool. It is knowing where to start. Large companies have teams to map their processes. Small teams are busy running the business.

What I would do this week

• Pick one repetitive process, like lead follow-up or client onboarding.
• Write down every step, and mark the steps that need human judgment.
• Let an agent handle only the rest, and review its work for two weeks before you trust it.

McKinsey's separate research on AI trust found that security and risk concerns are the top obstacle to scaling agentic AI. Starting small, with a person reviewing the output, answers both.

My takeaway: You do not need a big budget to close this gap. You need one well-chosen process. Automate the process, not the relationship.