Does using AI at work come down to your company's size and resources — or more to your role and whether anyone checks your output?

This explores whether firm size and resources decide who picks up AI, or whether other forces matter as much. The short answer from the corpus is that size matters at the front door, and other factors take over once you're inside. OpenAI's own telemetry from 1,764 companies shows ChatGPT Enterprise adoption concentrating in larger, R&D-intensive firms Who adopts enterprise AI first and how do they use it?. A separate study of online labor marketplaces finds that firms more exposed to AI replace contract workers with AI faster and more cheaply than other firms Do firms substitute labor for AI at different rates?. That second finding is the more surprising one. It suggests AI capability compounds inside a firm, so early adopters pull ahead instead of everyone moving forward at the same pace.

The same OpenAI data also shows a pattern inside firms. Marketing staff and early-career workers use the tool far more than executives and senior people. So "who uses AI" is partly a question about the company and partly a question about where you sit in it. Company rules matter too. A study of junior and senior engineers found that tool mandates, approved-tool lists and data policies set how much control people have over AI agents before personal preference comes into play Does personal preference shape how engineers use AI tools?. Being in a large firm can mean more access to AI, but also more limits on how you're allowed to use it.

Small firms show a different side of the story. In US court cases with suspected AI-generated errors, 90 percent involved solo practitioners or small law firms Do small law firms misuse AI more often than large ones?. That measures errors that were caught, not how often each kind of firm misuses AI. Still, a likely reading is that small practices use AI without the review layers that big firms have. Their problem isn't access to the tools. It's having the capacity to check the output. Developer surveys point the same way: 80 percent of developers use AI tools while their trust in its accuracy has fallen to 29 percent, mostly because of code that looks right but isn't Why do developers keep using AI tools they don't trust?.

Two more factors don't depend on firm size at all. People expect colleagues to see them as less competent when they use AI, so they hide it Do people fear judgment when they use AI at work?. That means official adoption figures may undercount real use. Benedict Evans argues that easier tools don't fix the deeper barrier: most workers don't see which of their own tasks could be automated, and enterprise rollout depends on decisions across departments, not technical ability Does easier tool-building actually solve enterprise adoption problems?.

The takeaway you might not expect: firm size predicts who signs the enterprise contract, but it's a weak predictor of who benefits. What separates firms is whether they can check AI output and build skill with it over time. That's why the corpus shows big firms pulling ahead on cost savings while small firms show up in the error records. The collection has little direct data on small and mid-sized businesses beyond the legal cases, so the middle of the firm-size range is mostly uncharted here.