AI use may be getting more acceptable in the marketing workflow, but local advertisers are drawing a line around what they expect to pay for it. New Borrell Associates data shows rising concerns about AI accuracy, disclosure, and paying human rates for machine-generated work.
Asked how comfortable they are with vendors, freelancers, or agencies using AI to complete marketing work on their behalf, local advertisers split into several camps, according to Borrell Associates. Just over half said they are generally comfortable with it, provided they are told AI is being used. Another 17% said they are completely comfortable, while 14% prefer human-only work, 10% oppose it outright, and 8% said they didn’t know.
Yet, that comfort comes with some growing skepticism. The share of local advertisers concerned that a vendor is using AI instead of doing the work themselves climbed to 35%, up from 18% a year earlier. 58% said they are concerned about paying full human rates for AI-generated work.
Accuracy is the biggest concern when advertisers’ outsourced partners use AI. Seventy-one percent said they worry the technology could introduce errors. When asked about AI more broadly, 72% cited the same concern that the technology could produce false information or hallucinations.
Borrell also found that among advertisers who have noticed changes in their own customers’ behavior, the most common shift is customers showing up with incorrect or misleading information they received from AI. Still, 73% said they have not noticed any change in customer behavior.
A similar divide showed up on the listener side of the business in a Harker Bos Group Crowd React study. The blind-listening test of 1,326 weekly radio listeners found AI-generated and human voiceovers scored nearly identically on trust, energy, and likability. But reactions changed sharply once listeners learned whether the voice they heard was AI-generated or human. Crowd React Media Founder Katie Miller framed that divide as a question of disclosure rather than audio quality. Borrell’s advertiser data points in much the same direction: AI itself may not be the dealbreaker. Knowing when it’s being used may be
