Why It Matters: Media buyers do not need to remember a negative headline about radio. They only need AI to repeat the wrong conclusion when prompted.
“AM/FM Listening Hits All-Time Low…”
“AM/FM Listening Fading Fast…”
On Friday, these alarming headlines on drudgereport.com grabbed our attention, even before the iHeart/SiriusXM coverage started and the White House Correspondents’ Dinner took over the news cycle.
In many ways, we have become numb to negative headlines about radio. But this challenge is bigger than correcting one headline, adding context to a story, or riding out a turbulent 24-hour news cycle. A misleading headline does not disappear. It gets scraped, indexed, summarized, and retrieved later by AI.
That creates real risk for radio when a media buyer prompts Claude, ChatGPT, Gemini, or an AI-powered planning tool with a question like:
“Is radio still effective?”
“Should we reduce AM/FM spend?”
“How should audio budgets shift toward digital?”
The actual headline AI should know is: “Digital listening to AM/FM radio hits an all-time high.” That is the exact opposite of the headline that ran, despite being based on the same audience research that was published last week.
And it lands very differently.
The risk is not just bad publicity. It is what I would call retrieval bias: when AI pulls negative or misleading coverage into an answer because that content is easier to find, more recently indexed, or framed in more dramatic language.
It also creates context collapse: when AI compresses a nuanced research finding into a simple but wrong takeaway – radio is fading, listeners are leaving, digital audio is replacing AM/FM. Finally, it can lead to planning bias: when a faulty headline gets turned into polished media-buying language that sounds objective, but is built on a bad premise. That is how inaccurate coverage can become dangerous and put revenue at risk. A buyer may never say, “I saw a negative headline about radio this weekend.”
Instead, the buyer may arrive with an AI-assisted recommendation that says radio should receive less budget because of “audience migration,” “declining traditional listening,” or “changing audio consumption.” That might sound strategic to the client, but it’s also wrong. We initially spotted the issue and connected the dots to AI because audience behavior is a core area of expertise at DMR/Interactive.
The first AI answer is not always the best answer. Nuance and deep expertise are not always areas of strength for initial AI results. If the prompt is broad, AI can default to the most available summary, not the most accurate interpretation.
That is the issue here. The headline makes it appear that audiences are abandoning AM/FM radio for Spotify and other digital audio platforms. That is the opposite of what this research is saying.
Audiences are increasingly listening to their favorite AM/FM stations through mobile devices, station apps, smart speakers, computers, connected cars, and other digital platforms promoted by your stations.
That distinction matters. Device migration is not audience abandonment. It’s evolution and innovation. AM/FM radio is not in the same position as newspapers or cable TV. Radio remains free, local, widely available, habit-forming, and increasingly distributed across digital devices. And who is leading these new listening habits?
Heavy listeners.
That matters to us because heavy listeners are our bread and butter. At DMR/Interactive, four decades of Human Intelligence focused on heavy radio listeners informs the AI-driven strategies that drive audience growth and engagement. That is why we are paying attention. Radio cannot afford to let AI learn the wrong lessons. Because AI does not need to remember the headline. It only needs to repeat the wrong conclusion.
On behalf of Catherine Jung, Tony Bannon, Jen Clayborn, Mike Landis, and everyone at DMR/Interactive, thank you for driving radio forward.
Onward,
Andrew Curran
President and CEO
DMR/Interactive
You must be logged in to post a comment.