Automotive dealers are increasingly adopting artificial intelligence into their business practices, but a new survey found that most are dissatisfied with general-purpose AI tools.

Lotlinx, a VIN performance platform, conducted the poll of about 215 dealership executives across the U.S. in March.

Eighty-four percent of respondents reported that they “often” or “almost always” fail to get what they need from generic AI tools. According to Lotlinx, researchers are calling the growing frustration with AI tools that don’t meet the modern dealership’s needs “Generic AI Fatigue.”

While 54% of dealers reported using generative AI tools weekly, they primarily use them for vehicle descriptions, marketing copy, social media and email templates. But there is growing interest in how AI can be used beyond such surface-level tasks. Specifically, dealers said they want an AI solution that understands their inventories and how their business operates.

Chatbots can be helpful, but dealers reportedly want more actionable AI that functions like a true assistant capable of more than just responding to prompts. Some of the capabilities they said they'd value are a “watch my store” monitoring feature and natural-language inventory queries.

When it comes to maximizing return on investment, more than 56% of dealers want a tool that can assess their inventory risks.

One area of concern that the survey uncovered lack of awareness among dealers of the data security risks associated with using AI tools. While 69% of respondents reported that they daily upload their data into generic AI tools, only 11% expressed concern about data security. According to Lotlinx, the lack of awareness could expose dealers to operational and competitive risk.

“For too long, dealers have been sold a one-size-fits-all view of AI, where generic tools are expected to solve highly specific business challenges,” said Lotlinx Chief Commercial Officer Randy Kobat.

“When nearly two-thirds of their peers say the tools available to them don't understand their own inventory and 84% say they fail to get useful results, that's not a technology problem, that's an industry problem."