At the AutoIndustry.ai Summit before ASOTU CON 2026, CarEdge Co-Founder Zach Shefska opened with a live test: 100 AI agents contacted 100 dealerships, asked for pricing, and started negotiating in real time.
The point was not the stunt. It was the warning.
Agentic AI is already entering retail automotive, and it is exposing the gaps between advertised prices, out-the-door quotes, add-ons, fees, response speed, and customer trust.
What Is Agentic AI in Automotive Retail?
Agentic AI refers to AI systems that can act on behalf of a customer or business, not just answer questions.
In this case, CarEdge deployed AI agents that could:
Shop dealership inventory
The agents contacted dealers about specific vehicles.
Request out-the-door pricing
They asked for full pricing with ZIP-code-specific taxes and fees.
Negotiate with dealerships
The agents were built to continue the conversation, not stop after one email.
Capture structured data
Shefska said each interaction can collect more than 160 data points, including pricing, fees, add-ons, response time, and transparency markers.
The Live Mystery Shop Results
Within the first hour, the AI agents generated:
83% dealer response rate
Shefska called the early response rate “fantastic,” especially from email and SMS outreach.
12 out-the-door quotes
The test showed dealers are capable of responding quickly, but the quality and clarity of those responses varied.
Early signs of pricing disconnect
One example showed a Honda Pilot listed online at $58,135, but the quote included add-ons like EQ Repel, Pro Pack, and a doc fee.
Why Pricing Transparency Is Becoming a Competitive Advantage
Shefska’s strongest point was simple: consumers reward honesty.
After CarEdge previously mystery-shopped 100 Ford dealers on similar F-150s, the lowest and highest out-the-door quotes had a $9,256 spread, excluding sales tax.
The surprising response from consumers was not just anger at bad actors. Many asked CarEdge to highlight honest dealers.
“People want to do business with people that they can trust.”
The Dealer Transparency Index: AI as Accountability Infrastructure
Shefska shared how he “vibe coded” the first version of the Dealer Transparency Index on a flight to NADA.
The goal: grade dealers based on verified pricing behavior, not just customer reviews.
One case study: $30,582 became $41,455.78
A vehicle advertised around $30,000 grew by more than $10,000 after certification, Toyota Care Plus, GAP, warranty, taxes, and fees were added.
Most dealers are not the problem
Shefska said CarEdge has graded more than 12,000 dealers across nearly 60,000 out-the-door quotes.
His reported breakdown:
60% earned an A
6% earned a D
3% earned an F
His conclusion: the bad actors are hurting the reputation of the whole industry.
What Dealers Should Do Now
1. Clean up your data
Shefska said the same rule from past tech shifts still applies: bad data creates bad outcomes.
“The cleanliness of your data.”
Inventory, pricing, CRM records, fees, add-ons, and availability need to be accurate before AI touches them.
2. Own your information
Dealers should not depend fully on third-party platforms to define their customer experience.
“You own your data. Build stuff with it.”
3. Prepare for AI-enabled retail as a third channel
Shefska framed the future as three sales paths:
Retail
Wholesale
AI-enabled retail
This does not mean every customer wants an AI-only car deal. It means some customers will use AI to search, compare, negotiate, and prepare before human handoff.
4. Keep the human parts human
Shefska does not believe car buying becomes fully AI-driven.
Test drives, trade evaluations, delivery, and relationship-building still belong to people.
AI should reduce friction in:
Search
Negotiation
Paperwork
F&I prep
Price comparison
The First AI-to-AI Car Deal Has Already Happened
Shefska cited Mark Miller Subaru as the first AI-to-AI car deal he is aware of.
The dealer-side AI negotiated with the customer-side AI, using inventory ingestion, pricing rules, and CRM handoff.
That is the shift: AI is not just helping customers. It is becoming part of dealership operations.
The Big Takeaway for Dealers
Transparency is no longer just a brand value. It is becoming measurable.
AI agents can compare prices, document fees, expose add-ons, track response times, and publish patterns at scale.
The stores that win will not be the ones that hide behind complexity. They will be the ones that make buying easier, clearer, and more efficient.
Shefska’s closing ask was direct:
“Help design it.”
The opportunity is to build an automotive retail future that is fair for customers, fair for operators, transparent for everyone, and more efficient for the business.
