A new report from CarEdge, a consumer-advocacy automotive platform shows how car dealers are changing advertised prices – red handed, and with receipts – exposing that the advertised price on a car dealer’s website may not be the price a buyer is actually offered.
AI agents created a digital trail spotlighting the difference at unprecedented scale … something individual consumers could never realistically do themselves.
During the investigation, AI Agents contacted scores or hundreds of dealerships simultaneously, requesting itemized out-the-door prices, comparing those quotes against advertised listings and preserving the conversations.
The same AI negotiator used the exact same buyer profile to negotiate with more than 100 dealers per vehicle, but results reveal striking discrepancies. Automotive AI analysts Zach and Ray Shefska reveal the investigation found as much as a $9,221 price spread on the exact same vehicle with quotes containing unrequested add-ons and documentation fees ranging from $0 to $1,094…and more than half of dealers contacted refused to disclose or commit to pricing over email or text.
A LIST OF VIDEO’S AND TEXT TRANSCRIPT RESULTS FROM THE EXPOSE HERE:
https://personal.filesanywhere.com/guest/fs?v=8e716786596470b1a497&C=50
The emerging story: AI isn’t just negotiating car prices anymore. It is becoming a real-time transparency tool capable of documenting the gap between what a dealership advertises online and what the same dealership tells a buyer when asked for the actual out-the-door price.
Among the findings and issues the experts can address:
When the advertised price changes once the buyer engages. CarEdge has documented examples in which the price presented online differs from what emerges during the actual dealer conversation, giving consumers a way to compare the listing with the real transaction.
AI can shop 100 dealerships so a consumer doesn’t have to. Instead of relying on one or two dealer quotes, an AI agent can systematically contact large numbers of sellers using the same buyer profile and request the same information, making price discrepancies far easier to identify.
Same car, same buyer, radically different prices. CarEdge’s RAV4 Hybrid experiment produced a $9,221 dealer-controlled spread among verified quotes, demonstrating how dramatically the ultimate price can vary depending on the dealership.
The add-ons buyers never asked for. Across CarEdge’s AI dealer experiments, 37% of quotes included add-on products that the hypothetical buyer had not requested.
Even the paperwork price can vary by more than $1,000. Documentation fees in the experiments ranged from $0 to $1,094 despite dealers performing essentially the same transaction paperwork.
Some dealers won’t put the price in writing. More than half of dealerships contacted in CarEdge’s experiments refused to provide pricing through email or text, raising another consumer question: why is it still so difficult to obtain a straightforward out-the-door price before entering a showroom?
AI creates a record of what was actually said. Automated outreach can preserve listings, quotes and dealer communications, making it easier to identify when advertised prices, required add-ons, financing conditions or other terms change during the sales process.
Federal regulators are already focused on the same pricing practices. In March, the FTC warned 97 dealership groups about deceptive pricing and specifically identified concerns including mandatory fees excluded from advertised prices, discounts unavailable to all consumers, required financing, mandatory add-ons and unavailable vehicles.
The scale is becoming significant. CarEdge reports that its AI system has conducted more than 146,000 dealer outreach sessions since July 2025 and generated more than $51 million in cumulative buyer savings, providing a growing dataset on what happens when consumers ask dealerships for actual transaction prices.
Dealer transparency can now be measured rather than reviewed. CarEdge’s Dealer Transparency Index analyzes verified out-the-door quotes and grades dealerships based on factors including fee transparency, add-on behavior and price accuracy, shifting the conversation from subjective online reviews to transaction-level pricing data.
The larger question is whether AI could become an unexpected new enforcement mechanism in the car-buying marketplace, not by replacing regulators, but by giving millions of consumers the ability to detect, document and challenge pricing discrepancies that historically occurred one customer at a time and largely out of public view.




