TO THE EDITOR:
Regarding "Dealerships look to build guardrails on AI amid dire warnings about technology," autonews.com, Sept. 30: Mark Hollmer's story got the question right. AlixPartners' Himanshu Khandelwal's tests for any artificial intelligence tool belong on the wall of every dealership: When is it asked to engage, does it act on its own, and who answers for it when it is wrong?
Every tool in the story talks to customers. That is where the risk lives, and the article's safeguards, a person approving and a manager auditing, are the right ones. But there is a second kind of AI the industry has barely touched, and it passes all three tests. It is asked to engage by the store's own people, every morning, with the store's own numbers. It does not act, it proposes, and the person who owns that lane decides. When it is wrong, the manager it coached catches it, because he was never cut out of the loop.
Alex Perdikis of Koons Motors Inc. hoped for employees who walk in to "a more structured, thoughtful day that's been set up by using AI." That does not require trusting a machine with your customer. The bar for any such tool is simple: It tells you what it sees, it leaves the decision with you, and it can show its work.
Dealers are right to be cautious about AI that faces the public. They should be just as demanding about AI that faces their people.
CEO
TelicX
Phoenix
TelicX is the company behind A.D.A.M. (Adaptive Dealer Advisory Mind), an AI advisor that reads a franchised dealership's own numbers and coaches dealership teams.
Read the letter at Automotive News. It answers "Dealerships look to build guardrails on AI amid dire warnings about technology," by Mark Hollmer, Automotive News, September 30, 2026.