[Small Business](https://bfcs.ai/industry/small-business/) · [Cheaper](https://bfcs.ai/cheaper/)

# Can AI Make Small Firms More Competitive?

The tools that once needed an in-house team now cost a monthly subscription, and the biggest measured gains go to the least experienced workers. That should favour small firms. So far the official data says large firms are pulling further ahead.

Updated **October 2026** · 6 min read · Evidence-graded · 8 sources

## The BFCS verdict

Every article scored on the same four dimensions

| Dimension | Verdict | Why |
| --- | --- | --- |
| Better | Partly · Moderate confidence | Biggest lift for the least experienced · Controlled studies show quality and output gains concentrated among novices, which suits small teams without specialists. The studies are of single tasks and single firms, not small businesses as a whole. |
| Faster | Yes · High confidence | Drafting and routine tasks compress · Randomised evidence shows large time savings on everyday writing work, and about a fifth of Australian SMEs in a government tracker cite faster decision-making as a benefit. |
| Cheaper | Partly · Moderate confidence | Cheap to start; savings slow to show · About a third of UK adopters report lower operating costs, but most have not yet seen revenue change, and a large Danish working paper finds no measurable effect on hours or earnings two years in. |
| Safer | Conditional · Governance-gated | Small firms carry the same data duties with less help · Off-the-shelf tools make it easy to paste customer data into systems the owner does not control. Safety depends on simple rules set before use, not on the tool. |

## The problem, in numbers

Small firms have always competed against scale. A large competitor can afford a marketing team, an analyst, a dedicated customer service desk and someone who reads the contracts. A five-person business does those jobs itself, after hours, or not at all. The promise of off-the-shelf AI is that a subscription now buys a passable version of each.

The official data says that promise has not yet closed the gap. In Australia, around 12% of businesses used AI in 2024-25, but 35% of large businesses did, up from 9% in 2021-22. (Source 1) In the United States, 37% of firms with 250 or more employees reported using AI by May 2026, against less than 20% of firms with four or fewer, and use among firms under 20 employees did not change significantly over the preceding five months. (Source 3) The UK picture is the same shape: 36% of large businesses against 14% of micro businesses. (Source 4)

The National AI Centre's tracker reports a much higher figure, 41% of SMEs "adopting" AI, from responses gathered in late 2024 and released in June 2025. (Source 2) The difference is likely mostly definition: a monthly sample of SMEs asked about adoption is not the same instrument as the ABS's business survey of AI use. The tracker covers only SMEs, so it cannot show the size gap; the ABS figures do.

### The evidence, in four numbers

35% vs 12% · large businesses using AI vs all businesses · Australia, 2024-25 (ABS Characteristics of Australian Business)

19% vs 4% · AI use, innovating vs non-innovating small firms · Australia, 2024-25 (ABS); innovation, not size alone, tracks use

34% · UK AI users reporting lower operating costs · DSIT AI Adoption Research, fieldwork Feb to May 2025

77% · UK AI users yet to see any revenue change · same survey; the competitive payoff is still mostly ahead

## What improves, and why

The case for small firms rests on who gains most. In a large field study of customer support agents, AI assistance raised issues resolved per hour by 15% on average, with the gains concentrated among less experienced and lower-skilled workers; the working paper version put the novice gain at 34%, with minimal effect on the most experienced. (Source 5) In a randomised experiment on professional writing tasks, ChatGPT cut average time by 40% and raised quality by 18%. (Source 6)

That pattern matters for a small business. A large firm already has the specialist who writes the tender, drafts the policy or answers the hard customer email. A small firm usually has a generalist doing it for the first time. If AI lifts the floor more than the ceiling, the firm with the lower floor has more to gain per person. The economics of access have also changed. Capabilities that once justified a hire or an agency, such as first-draft marketing copy, a summary of a long contract, a spreadsheet formula or a translated product page, now come bundled in general-purpose tools priced per user. For a large firm that is a marginal saving on work it already does well. For a small firm it can be the difference between doing the task adequately and not doing it at all, which is where competitiveness is actually decided.

In the National AI Centre's tracker, 22% of surveyed businesses report faster decision-making and 18% improved productivity, though the release does not say whether those shares are of all SMEs or only of adopters. (Source 2)

> The tools are cheap enough for anyone. The capability to use them well is not, and that is where the size gap actually sits.

## What the evidence actually says

The cost side is real but early. Among UK businesses already using AI, 34% report reduced production or operating costs. Yet only 12% report higher revenue, and 77% have not yet seen any change in revenue at all. (Source 4) A cheaper process is not the same as a stronger competitive position until it turns into price, capacity or customers.

The most rigorous economy-wide check is sobering. Linking adoption surveys to Danish administrative records, Humlum and Vestergaard estimate precise null effects on earnings and recorded hours, ruling out effects larger than 2% two years after ChatGPT's launch, even though workers report productivity benefits. (Source 7) This is a working paper, but it fits the UK revenue data: individual tasks speed up well before anything shows in the accounts.

The most useful Australian finding is about capability rather than size. Small businesses that were innovation-active used AI at 19%, almost five times the 4% rate of small businesses that did no innovation. (Source 1) By our reading, that puts innovating small firms close to the 22% rate of medium businesses overall, and suggests the divide that matters is between firms that already change how they work and firms that do not.

## Where it falls short

-   **The gap looks to be widening, not closing.** In the US, use rose among firms with 20 or more employees but not among smaller ones; in Australia, large-firm use nearly quadrupled since 2021-22. (Source 1) (Source 3) If AI helps whoever uses it, slower adoption by small firms hands large competitors a further edge.
    
-   **Task gains are not firm gains.** The strongest studies measure single tasks or single companies. Neither tells an owner whether the business will win more work.
    
-   **Surveys disagree on the level.** Rates of 12%, 14% and 41% describe different questions and samples. Treat any single headline adoption figure with care.
    
-   **Self-reported benefits.** Cost and productivity gains in adoption surveys are what owners believe, not what auditors measured.
    

## Where governance decides the outcome

A small business using a public chatbot carries the same data duties as a large one, usually without a privacy officer or IT team. The Australian privacy regulator recommends, as best practice, that organisations do not enter personal information, and particularly sensitive information, into publicly available generative AI tools. (Source 8) For a small firm that means a short written rule: what may be pasted in, which tools are approved, who checks output before it reaches a customer, and what happens when the tool is wrong. Safer is not a feature of the subscription. It is the owner's decision about how it is used.

## How to think about adoption

Start with the tasks where the evidence is strongest: drafting, summarising and routine customer responses, where time savings are large and well measured. Give the tool to the least experienced person on those tasks first, because that is where the lift is biggest. Measure the time saved, then decide deliberately what to do with it: lower prices, take on more work, or answer customers faster. Without that last step, the saving stays inside the task and never reaches the competitive position. And treat adoption as part of a wider habit of changing how the business works, because in the Australian data that habit, more than headcount, is what tracks who uses AI at all.

## How we read this evidence

BFCS.ai does not fabricate figures. Adoption surveys define "using AI" differently, so their rates are not directly comparable; we report each on its own terms. The Danish labour-market study is a working paper and is labelled as such.

## Sources

1.  Australian Bureau of Statistics, "Business adoption of Artificial Intelligence accelerates in 2024-25", media release on Characteristics of Australian Business 2024-25, 25 June 2026. [abs.gov.au](https://www.abs.gov.au/media-centre/media-releases/business-adoption-artificial-intelligence-accelerates-2024-25)
    
    **What we took from it:** Around 12% of businesses used AI; 35% of large and 22% of medium businesses; innovation-active small businesses 19% vs 4% for non-innovating small businesses.
    
2.  Minister for Industry and Innovation, "Australian AI Adoption Tracker report shows business harnessing AI", media release, June 2025 (tracker published 4 June 2025; responses gathered October to December 2024). [minister.industry.gov.au](https://www.minister.industry.gov.au/ministers/timayres/media-releases/australian-ai-adoption-tracker-report-shows-business-harnessing-ai)
    
    **What we took from it:** National AI Centre tracker (with Fifth Quadrant, 400+ businesses surveyed monthly): 41% of SMEs adopting AI; 22% of businesses report faster decision-making; 18% improved productivity; 21% unaware how to use AI. The release does not say whether the benefit shares are of all SMEs or only adopters. Ministerial page timed out on checking; figures confirmed via trade-press coverage and the tracker's Q4 2024 summary.
    
3.  U.S. Census Bureau, Grundy, Breaux and Khatiwoda, "Large Firms With at Least 20 Employees Biggest AI Users", Census Bureau story on the Business Trends and Outlook Survey, May 2026. [census.gov](https://www.census.gov/library/stories/2026/05/ai-use-businesses.html)
    
    **What we took from it:** 37% of firms with 250+ employees used AI; less than 20% of firms with four or fewer employees; no significant change for firms under 20 employees between December 2025 and May 2026.
    
4.  UK Department for Science, Innovation and Technology (IFF Research and Technopolis), "AI Adoption Research", fieldwork 12 February to 2 May 2025, page updated 13 February 2026. [gov.uk](https://www.gov.uk/government/publications/ai-adoption-research/ai-adoption-research)
    
    **What we took from it:** 3,500 businesses with 5+ employees: 36% of large, 23% of medium and 14% of micro (5 to 9 employees) businesses use AI. Among the 700 AI users, 34% report reduced production or operating costs, 12% higher revenue and 77% no revenue change yet.
    
5.  Brynjolfsson, E., Li, D. and Raymond, L. R., "Generative AI at Work", The Quarterly Journal of Economics 140(2), 889 to 942 (2025); NBER Working Paper 31161. [nber.org](https://www.nber.org/papers/w31161)
    
    **What we took from it:** Published version: among 5,172 customer support agents, AI assistance raised issues resolved per hour by 15% on average, and less experienced and lower-skilled workers improved both speed and quality. The working paper (5,179 agents) reported 14% on average, 34% for novice and low-skilled workers and minimal impact on experienced and highly skilled workers.
    
6.  Noy, S. and Zhang, W., "Experimental evidence on the productivity effects of generative artificial intelligence", Science, July 2023, doi 10.1126/science.adh2586. [scale.stanford.edu](https://scale.stanford.edu/publications/experimental-evidence-productivity-effects-generative-artificial-intelligence)
    
    **What we took from it:** Preregistered experiment with 453 college-educated professionals: ChatGPT cut average time on writing tasks by 40% and raised output quality by 18%.
    
7.  Humlum, A. and Vestergaard, E., "Still Waters, Rapid Currents: Early Labor Market Transformation under Generative AI", NBER Working Paper 33777, May 2025, revised March 2026 (not yet peer-reviewed). [nber.org](https://www.nber.org/papers/w33777)
    
    **What we took from it:** Danish survey and administrative data: precise null effects on earnings and recorded hours, ruling out effects larger than 2% two years after ChatGPT launched.
    
8.  Office of the Australian Information Commissioner, "Guidance on privacy and the use of commercially available AI products", 21 October 2024, updated 17 January 2025. [oaic.gov.au](https://www.oaic.gov.au/privacy/privacy-guidance-for-organisations-and-government-agencies/guidance-on-privacy-and-the-use-of-commercially-available-ai-products)
    
    **What we took from it:** Recommends, as best practice, that organisations do not enter personal information, particularly sensitive information, into publicly available generative AI tools.
    

Putting this to work

## Working through this in your own organisation?

BFCS.ai is published by Inhouse CX. We help small and mid-sized organisations decide where AI earns its place, and put the governance in place that makes it safe to use.

[Talk to Inhouse CX →](https://inhouse.cx/engage/)

Continue the thread

- [Can AI Speed Up Admin for Owners?](https://bfcs.ai/articles/can-ai-speed-up-admin-for-owners/): More in Small Business · SMALL BUSINESS · FASTER · 6 min · Read →
- [Can AI Improve Small-Business Customer Service?](https://bfcs.ai/articles/can-ai-improve-small-business-customer-service/): More in Small Business · SMALL BUSINESS · BETTER · 6 min · Read →
- [Can AI Reduce Delivery Costs?](https://bfcs.ai/articles/can-ai-reduce-delivery-costs/): Also Cheaper · PROFESSIONAL SERVICES · CHEAPER · 6 min · Read →

---

Canonical page: https://bfcs.ai/articles/can-ai-make-small-firms-more-competitive/
This is the Markdown version of the page, generated from its HTML at build time.
Site guide for agents: https://bfcs.ai/llms.txt · Every page: https://bfcs.ai/sitemap.xml