29 September 2026

The AI Divide

The gap regulated small firms can't afford to ignore

This essay sits primarily under the Can we do AI? question of the practice’s methodology: the literacy gap that already separates firms moving past surface-level AI use from those still circling it.

An opinion piece by Michael Borck

There is a conversation in professional services right now about how AI will transform legal work, clinical practice, accounting, and financial advice. It is important. It also has a blind spot.

The divide is not coming. It is already here. It operates at two levels: the practitioner sitting at their desk, and the firm that desk sits inside.

The practitioner-level divide

In any given week, some practitioners are working with GPT-4-class models, Claude Pro, Microsoft 365 Copilot, and an assortment of vendor-embedded AI features. Others are working with whatever free tier they can pull up on a personal phone.

This is not hypothetical. It is happening now in suburban legal practices, regional clinics, and accounting firms across Australia.

Practitioners who have access to enterprise-tier AI (properly licensed, properly governed, properly integrated) are operating with fundamentally different tools than those who do not. Better drafting support. More sophisticated research. Faster summarisation. The gap between free and paid is widening, not narrowing, and the gap between properly governed and ad-hoc is wider still.

We talk about AI as a great equaliser for small firms. We should stop. AI is currently a great amplifier: of advantage for those who can deploy it well, and of disadvantage for those who cannot.

The firm-level divide

Pull the lens back to firms, and a second fracture appears.

A top-tier law firm with an in-house technology team operates in a different reality from a five-partner suburban practice. A national health network with a clinical informatics function operates in a different reality from a sole-practitioner GP. A Big Four firm piloting AI agents operates in a different reality from a regional accounting practice with one senior partner and a part-time bookkeeper.

The difference is not interest. The smaller firms I work with are often more curious about AI than the larger ones. The difference is structural: capacity to evaluate vendors, capacity to negotiate enterprise terms, capacity to write meaningful policies, capacity to train staff, capacity to absorb the cost of getting it wrong once.

Well-resourced firms are already running pilots, building internal tools, and producing the case studies that will define market expectations. They have runway to experiment and to fail.

Under-resourced firms face a different calculus. When margin is thin and any misstep invites a complaint or a regulator letter, innovation looks like risk. The rational response is retrenchment: blanket bans, vague policies, defensive procurement decisions, and a quiet hope that this will pass. It will not.

The distance between the two grows with every quarter. Along fault lines of capital, geography, and structural advantage that were entrenched long before generative AI arrived.

We have seen this before

We saw it with the internet. We saw it with practice management systems. We saw it with cloud accounting. Same script every time: early adoption by the well-resourced, defensive reaction by the under-resourced, a widening gap dressed up as digital transformation.

The difference this time is speed. Previous cycles gave firms years to catch up. AI capabilities advance quarterly. Practical implications shift monthly. Standing still means falling behind at an accelerating rate, and what falls behind is not just the technology: it is the firm’s competitive position, its ability to attract and retain staff who want to work somewhere that is not stuck, and its standing with the clients who can already see what their bigger options are doing.

What it takes to close the gap

The good news is that small firms are not powerless. The advantages of being small (proximity to clients, faster decisions, less internal politics, a smaller surface area to govern) are real, and they cut in your favour if you let them.

In practice:

  1. Skip the enterprise theatre. Small firms do not need a hundred-page AI policy. You need a short, specific one (written for this firm, these obligations, these tools) that staff actually read.
  2. Buy what you already pay for. Most small firms in Australia are on Microsoft 365 or Google Workspace. The AI inside those platforms is enterprise-class and largely already in your existing data agreements. Use it before you go shopping.
  3. Train people, not policies. A three-hour partner session on how to evaluate AI output is worth more than a thirty-page document nobody reads.
  4. Decide whether you need to act before deciding what to buy. The vendor pitch always has urgency built in. Your situation may not.
  5. Centre the voices closest to the work. The conversation about AI in professional services is dominated by people with the largest budgets and the loudest microphones. They are not your firm. The most useful peers are other small firms who have done this honestly. Find them.

The stakes

The promise of AI for small regulated firms is real. Better drafting. Faster research. More time for clients. More room for principals to be senior again. But these will remain marketing copy unless the gap is named honestly.

There is also an agency dimension to the divide that the bigger firms cannot solve for you. A small firm that lets a vendor (or a sector body, or a peer) do its thinking for it gives up the one advantage being small still confers: the ability to choose, fast, on your own terms. Closing the gap is not just about adopting tools. It is about staying the kind of firm that makes its own decisions.

The AI divide is not a side issue for the small-firm sector. It is the issue. Every conversation about AI in your practice should start with a question: who gets left behind, and how do we make sure it is not us?

If you cannot answer that honestly, you are not transforming the practice. You are just making the gap faster.

About this work

borck.consulting is built around a simple wager: that regulated Australian small firms can adopt AI well, on their terms, without copying the playbook of firms ten times their size, if they get the framing right early. Engagements begin with the AI Readiness Diagnostic.

References