AI agents for car dealerships

A dealership's website gets questions at every stage of owning a car: is this one still here, what is the real price, what is my trade worth, can I drive it tomorrow, and, years later, when is my next service. Many arrive after hours, from shoppers comparing three dealers at once. An AI agent can answer from your vehicle and service pages, book test drives and service visits, and hand a well-qualified shopper to your sales team. Prices, payments and credit carry legal rules, and those decide what the agent may say about money.

Who visits a dealership website

• Shoppers comparing specific vehicles. Availability, price and fees, history, features, and how to see the car.
• Trade-in owners. What their current car is worth and how the appraisal works.
• Finance shoppers. Monthly payments, preapproval, leasing versus buying, what credit they need.
• Service customers. The largest group of repeat visitors: prices, appointments, loaner cars, recalls, warranty work.
• Business and fleet buyers. Several vehicles, upfitting, invoicing and fleet programs.

Which ASDK agents fit a dealership

ASDK is in early access. All agents answer from one knowledge base built from your website: vehicle pages, specials, service menus and policies.

Sales: from vehicle question to test drive (early access)

Example flow: "Is the blue RAV4 Hybrid still available? Can I drive it tomorrow after 5?" Sales answers from the vehicle page, says availability is confirmed by the sales team, then asks what a salesperson needs: timeline, trade-in yes or no, and whether the shopper plans to finance, lease or pay cash. It books the test drive through Cal.com, Calendly or Google Calendar and sends the lead to your team with the vehicle and the answers; native CRM sync is coming soon. Sales is included from the Pro plan.

Fleet and business buyers get a different path. When a visitor asks about six work vans with shelving, Sales asks about quantities, body and upfit needs, timing and whether the company has a fleet account, then books a call with your commercial or fleet manager instead of the floor team. The same rules apply: no prices beyond what you publish, and no financing terms in the chat.

Chat: vehicles, service and policies (early access)

Example flow: "How much is an oil change for a 2021 Camry, and do you have loaner cars?" Chat answers from your service menu and policies, with links, and offers to book. Asked "is the price negotiable?", it repeats your published pricing policy and nothing more.

Search: inventory and service answers (coming soon)

When Search ships, a query such as "hybrid SUVs under $30,000 with all-wheel drive" will get a short answer with links to matching vehicle pages, instead of a filter page the shopper has to work through.

Agent-Ready: readable for AI assistants (early access)

Agent-Ready hosts an llms.txt file that points AI systems to your inventory, service department and how to book.

Guardrails for dealerships

Prices as published, fees included. The agent repeats the advertised price and the fees your pages list, and links the vehicle page. It does not calculate an out-the-door figure that depends on taxes, registration and the buyer's address; the sales team does.

No payment math in chat. Under Regulation Z, an advertisement that states the amount of a payment, a down payment, the number of payments or the finance charge must also disclose the down payment, the repayment terms and the APR. Leases have similar rules under Regulation M. A chat answer such as "about $450 a month with $3,000 down" can trigger them. Instruct the agent not to state payments or rates and to link your finance page or payment calculator, where the full disclosures appear.

Credit applications stay in your secure systems. The FTC's dealer FAQ says dealers that finance, or help arrange financing for, consumers are financial institutions under the Safeguards Rule, which requires a written information security program. Social Security numbers, income and employment details never go into the chat; the agent links your secure credit application.

Trade-in values come from an appraisal. The agent explains how your appraisal works and books one. It does not guess a number.

Recalls: send owners to the official lookup. For "does my car have an open recall?", the agent links NHTSA's recall lookup, where owners can search by VIN, and offers a service appointment for recall work.

Say it is AI. ASDK always tells visitors they are talking to an AI, and a handoff to a person in sales or service is one step away. Negotiation always happens with a person.

How AI assistants research dealerships

Shoppers now ask assistants for "a dealer near Columbus with a certified pre-owned hybrid SUV under $30,000". Assistants and agentic browsers read vehicle pages, so make them readable as text, not only photos:

• Price, fees, mileage, VIN, trim, features and certification status on each vehicle page.
• Your pricing, delivery and return policies.
• Service menu prices, hours, loaner policy and how to book.
• Brands you sell and service, locations and languages spoken.

AutoDealer structured data can describe the dealership, its departments and their opening hours. The free Agent Readiness Checker shows what AI agents read on your site today, and Agent-Ready keeps llms.txt current as inventory turns over.

Where to start

Agree with your finance and compliance leads on what the agent never says: payments, rates, approvals, trade-in values. Then start Sales on vehicle pages and Chat on service pages, and read the first week of conversations. Pricing shows which plan includes each agent.