Dealership AI in 2026: Why the Big Groups Are Pulling Ahead — and How Independents Keep Pace

On June 29, Automotive News reported that the six largest publicly traded dealership groups are investing heavily in artificial intelligence across their operations, with Group 1 Automotive alone running more than 50 AI projects. The framing of the piece — how the public groups are "raising the bar" — is the part worth sitting with. When the biggest, best-capitalized retailers in the business all move the same direction at once, the bar they raise becomes the one every other store gets measured against: by shoppers comparing response times, by lenders watching how fast deals fund, and by the manufacturers grading customer satisfaction.

For a single-rooftop store or a small group, that can land as either a threat or as noise. It is neither. The more useful reading is that the public groups have started converting a belief almost everyone in the industry already holds into a short list of shipped, measured use cases. That conversion — not the size of the check — is the real gap. And for once, it is closable without a national balance sheet.

• Myth: the big groups are pulling ahead on AI because they can outspend everyone else.
• Data: 81% of dealers already say AI is here to stay, yet only about 15% have embedded it into daily workflows (Cox Automotive AI Readiness Study, 2025). The advantage the public groups are building is focus, not budget.

What "raising the bar" actually means

Strip away the vendor noise and the industry is in an unusual spot: nearly everyone agrees on the destination, and almost no one has arrived. Cox Automotive surveyed 537 franchise dealership leaders for its 2025 AI Readiness in Auto Retail Study and found broad conviction paired with thin execution. Most dealers — 81% — believe AI is permanent, and 63% say investing now is critical for long-term success. But 60% describe themselves as only beginning to test AI tools, and just under 15% have actually embedded AI into workflows and day-to-day decisions.

That is the whole story in two numbers. Belief is near-universal; operational adoption is rare. What the public groups have that a typical store does not is the scale to stand up dedicated teams and run dozens of experiments at once — Group 1's reported 50-plus projects is a portfolio, not a single tool. An independent cannot and should not try to run 50 projects. It does not need to. It needs three that work.

The concerns column matters as much as the adoption column. Nearly three-quarters of dealers worry about AI accuracy, and two-thirds want more training. Those are not reasons to wait; they are the specification for how to adopt — narrowly, with guardrails, and with staff who understand exactly what the system does and does not do.

Why the gap matters more in this market

The backdrop makes execution urgent rather than optional. Cox Automotive's late-June forecast put the June new-vehicle sales pace near a 16.1 million SAAR — steady, but with first-half 2026 volume tracking about 3.6% below the same period in 2025. The full-year forecast sits at 15.8 million units, down roughly 2.9% from last year, and affordability remains the ceiling — one Cox attributes more to elevated interest rates and stretched household budgets than to sticker prices alone.

A flat market changes where growth comes from. When the overall pie is not expanding, incremental sales and service revenue come from capturing demand that already exists — the lead that arrived at 9 p.m., the caller who could not get through to the service drive, the shopper who messaged three stores and bought from whichever answered first. Those are precisely the tasks today's AI handles well, and precisely where a store that believes in AI but has not deployed it leaks money every day.

Where AI is actually earning its keep

It is worth being specific about what dealers who have adopted AI actually use it for, because it is far more mundane — and far more dependable — than the agentic-commerce headlines suggest. In the Cox study, the leading AI use case was engaging customers 24/7 with automated text, chat, and email (52%), followed by personalized messages (48%) and predicting which shoppers are ready to buy (39%).

Notice that the top use case is not autonomous negotiation or self-driving F&I. It is response — being there instantly, at any hour, in the channel the customer chose. That maps directly onto the two leaks every store already knows about: missed and abandoned phone calls and slow lead response. Fixing those does not require a 50-project portfolio. It requires one dependable system that answers, qualifies, and books — then hands a warm, logged interaction to a person.

The independent's playbook

The public groups did not out-think smaller dealers; they out-focused them. Four principles from their approach translate to any store, regardless of size.

Pick three use cases, not fifty. Start where the return is already proven and the failure mode is cheap: after-hours and overflow phone coverage, instant lead response, and service scheduling. These are high-volume, repetitive, and easy to measure. Resist the urge to boil the ocean.

Connect it; do not bolt it on. Cox's Lori Wittman cautioned that AI "can't be just another point solution in a dealer's tech stack." The average dealership already runs more than 40 software systems. An AI tool that cannot see the CRM, the DMS, and live inventory will produce confident, wrong answers — the exact accuracy problem 74% of dealers name. Integration is not a nice-to-have; it is the line between help and liability.

Put guardrails and training around it. The two-thirds of dealers who want more training are right to. The stores that succeed treat AI like a new hire: a defined scope, a clear path to escalate to a human, and regular review of what it said. That is also the honest answer for compliance-minded managers — a narrow, supervised system is far easier to govern than a sprawling one.

Measure it like the groups do. The public groups run projects, not purchases, which means every deployment has a baseline and a number attached. Track answer rate, speed to first response, appointments set, and show rate before and after. If a use case cannot show a number in 60 to 90 days, cut it and move on.

What good looks like in twelve months

A realistic target for a small store or group is not 50 projects. It is a phone answered every time, a lead that gets a real response in seconds rather than hours, and a service schedule that fills itself overnight — each one wired into the systems you already run, each one with a before-and-after number a GM can defend in a manager meeting. That bar is within reach, and it happens to be the same bar the public groups are quietly clearing one project at a time.

The widening gap in dealership AI is real, but it is not primarily a spending gap. It is an execution gap — and execution is the one advantage a focused, well-run store can still win on.

If you are deciding where a single, well-integrated AI use case would pay for itself fastest, Carbuki builds AI voice agents that answer every sales and service call, respond to leads instantly, and book appointments straight into the tools a dealership already uses — a practical first project rather than a fiftieth.

Sources

• Automotive News, "Dealership AI: How the public groups are raising the bar" (June 29, 2026): autonews.com
• Cox Automotive, "Automotive Dealers Are Ready for AI to Deliver Outcomes and Skip the Hype" — AI Readiness in Auto Retail Study (Oct. 28, 2025): coxautoinc.com
• Cox Automotive, "New-Vehicle Sales Pace Holds Strong Through First Half of 2026" (June 24, 2026): coxautoinc.com