Three Years Into Enterprise AI: Where Most Firms Actually Stand

Three years after ChatGPT launched, I've worked with enough professional services firms to see the full spectrum. Some are genuinely advanced. Most are somewhere in the middle. Some haven't started. Let me be honest about where the bell curve sits.

The Distribution

If I had to segment firms by AI maturity in January 2026:

• 5%: Advanced. AI integrated into core workflows. Measured ROI. Serious governance. Competitive advantage visible.
• 20%: Intermediate. One or two AI projects working. Some staff using AI routinely. Investment ongoing but results unclear.
• 40%: Experimenting. Some tools deployed. Uneven usage. No clear strategy. Hope mixed with uncertainty.
• 25%: Dabbling. Partners using ChatGPT personally. No firm investment. No strategy. "We're watching."
• 10%: Not engaged. AI is not a priority. "Let us know when it's mature."

What Advanced Looks Like

The 5% who are genuinely advanced have these characteristics:

• Clear CEO/managing partner commitment and investment
• Dedicated AI leadership (person or team)
• AI integrated into at least two core processes
• Documented governance and compliance approach
• Measured ROI on specific projects
• 20%+ of staff using AI in their daily work
• Data quality and integration projects completed

These firms see real productivity gains: 15–30% for affected workflows. They have competitive advantage, at least for now.

What Intermediate Looks Like

The 20% in the intermediate range:

• Launched one or two specific AI projects
• See results from the projects (usually 15–25% productivity gain)
• Some governance in place, but incomplete
• 5–15% of staff using AI as part of their workflow
• Haven't yet figured out how to scale beyond the pilot
• Uncertain about next steps

These firms are on an interesting cusp. They're further along than the majority, but they haven't yet figured out if this is durable advantage or temporary novelty. Many will scale successfully. Some will stall.

What Experimenting Looks Like

The 40% in the experimenting zone:

• Multiple tools deployed (ChatGPT, Claude, some specialized tools)
• Usage is uneven—some people use AI daily, others have never tried it
• No clear measurement of impact
• Concerns about governance, but no formal policy
• Budget is scattered (tool subscriptions, but no strategic investment)
• No clear strategy or roadmap
• Leadership is "interested" but not committed

This group is where most firms are. They're not stuck, but they're not moving confidently either. The uncertainty is real.

What Dabbling Looks Like

The 25% dabbling:

• Partners and some staff have personal ChatGPT/Claude accounts
• Using AI for personal productivity, not firm strategy
• No firm-wide tools or investment
• No governance, no measurement, no strategy
• Attitude: "We're keeping an eye on it."

This group is waiting. For clarity. For better tools. For costs to drop. For competitors to move first.

The Uncomfortable Truth

Most firms (65%) are either experimenting or dabbling. They're not sure if AI is important. They're not moving fast, but they're not comfortable standing still. That's a tough place to be.

For the 5% that are advanced: the competitive gap is widening. It's not huge yet, but visible. A firm that's deployed AI well will deliver faster, at better quality, or at lower cost. Over time, that compounds.

For the 40% in the middle: the good news is you can still catch up. The gap isn't unbridgeable yet. But the window is closing. If you're still "experimenting" in 2027, you're behind. If you're advanced in 2027, you're ahead.

Why Are So Many Firms Stuck in Experimenting?

It's not lack of intelligence or laziness. It's legitimate challenges:

Uncertainty about governance. Professional services firms are regulated or quasi-regulated. Using AI in client work raises questions. There's no clear guidance. Firms are rightfully cautious.

Unclear ROI. Productivity improvements are real but often small (10–20%). That's good, but not transformational. It's hard to justify investment for modest gains.

Integration difficulty. Most AI tools don't integrate cleanly with legacy systems. You need custom work. That's expensive and slow.

Change resistance. People have worked the same way for years. Change is friction. Even if the tool is 30% faster, if adoption is 20%, the overall impact is 6%. Hard to justify.

Leadership uncertainty. When the leader is uncertain, the organization follows. Many managing partners don't know enough about AI to lead confidently. They delegate, but delegation without clarity doesn't work.

What Shifts a Firm From Experimenting to Intermediate or Advanced?

From what I've seen, it usually takes one of three things:

1. A burning platform. A competitor moves fast. A major client asks about AI capability. A partner retires and a new one brings AI vision. Something creates urgency.

2. A clear early win. One AI project works really well. It's measured, it's clear, it's proven. That creates momentum for the next one.

3. Leadership commitment and resources. A managing partner or executive team decides: "This matters. We're investing." They put a person on it. They allocate budget. They follow up.

Firms that do one of these three things start moving. Firms that don't stay stuck.

Where I See This Going

By end of 2026, I expect the distribution to shift:

• Advanced: 10% (doubled)
• Intermediate: 30% (jumped from 20%)
• Experimenting: 35% (down from 40%)
• Dabbling: 20% (down from 25%)
• Not engaged: 5% (same)

The momentum will be toward action, but most firms will still be uncertain.

If You're in the Experimenting Zone

Here's my advice: Pick something specific. Invest in it. Measure it. Scale it or kill it. Then pick the next thing. Momentum comes from completing cycles, not from thinking about what you might do.

This is not about moving fast. It's about moving with purpose and clarity. One good decision, well-executed, beats ten half-baked experiments.