How OOH is a Trust Medium in the AI Era

There is a question that nobody in advertising wants to answer honestly, because answering it honestly would require admitting that the thing they have spent the last fifteen years building might be, in the most fundamental and non-metaphorical sense, fake.

The question is this: Who is seeing your ads?

Not “which demographic.” Not “what cohort.” Not “adults 25-54 with household incomes above $75,000 who have recently expressed purchase intent in the athletic footwear category.” Who. As in, is it a person. A human person, with a body and a credit card and a commute and a reason to care about whatever you are selling. A person who could, at least theoretically, walk into a store and buy something. A person who exists.

This seems like it should be an easy question. It is not.

I. The Fifty-One Percent Problem, or: Who Is Seeing Your Ads and Are They Even Alive

In 2024, for the first time in a decade, automated bots accounted for more than half of all internet traffic. Fifty-one percent. That number comes from Imperva’s Bad Bot Report, which has been tracking this for twelve years, and which reads less like a cybersecurity document and more like the opening chapter of a novel about the slow, unremarkable death of something important. The kind of novel that gets optioned but never produced because the ending is too bleak and there is no protagonist. Thirty-seven percent of all web traffic is now classified as malicious. Not neutral. Not ambiguous. Malicious. Up from thirty-two percent the year before. Rising every year for six consecutive years. With no evidence, structural or otherwise, that it will stop.

Let me put that differently. If you showed up to a party and someone told you that more than a third of the guests were pickpockets, you would leave. You would not stand around the kitchen arguing about whether the guacamole justified the risk. And yet the digital advertising industry, which is a $400 billion party in the United States alone, has essentially decided that the guacamole is incredible.

Here is what that means if you buy or sell advertising for a living. More than a third of the internet is now a hostile simulation of itself. And it is getting worse, because AI has made it trivially easy to generate the kind of content that used to require human effort, human judgment, human taste. The open web is filling up with what the industry has started calling “AI slop” - machine-generated content trained on machine-generated content, in a feedback loop that degrades quality with each iteration. Researchers have a clinical term for this. They call it model collapse. It sounds like something a structural engineer would say about a parking garage. It should sound alarming in a way that makes you rethink your entire media plan.

Meanwhile, you are still buying impressions there.

The Association of National Advertisers ran the numbers. They found that only thirty-six cents of every dollar that enters a demand-side platform actually reaches the consumer. Thirty-six cents. I want to sit with that for a moment because it is genuinely staggering. If you gave someone a dollar and asked them to buy you a candy bar, and they came back with thirty-six cents worth of candy bar and a vague explanation about intermediary fees and supply chain complexity, you would never give that person a dollar again. And yet the programmatic supply chain processed roughly $180 billion in U.S. display spending last year. The ANA estimated $26.8 billion in waste in 2025 alone, up thirty-four percent from two years earlier. The World Federation of Advertisers confirmed it: for every $1,000 invested in programmatic advertising, only $439 reaches consumers as viewable impressions on valuable inventory. Fifty-six percent disappears.

The IAB found that fifty-seven percent of marketers now say they are less confident in data from programmatic platforms. Which is a remarkable thing to admit, given that programmatic platforms are where most of them are spending most of their money. It is like saying you are less confident in the structural integrity of the building you are currently standing in. The polite response is to nod. The correct response is to walk outside.

So let me restate the question. Who is seeing your ads?

And now let me ask a different one: What if the answer doesn’t have to be this depressing?

II. The Binary Is Broken and Nobody Wants to Admit It

For as long as anyone in media can remember, the fundamental organizing principle of the industry has been analog versus digital. Old versus new. Dumb versus smart. The billboard versus the banner ad. This framework has been so persistent and so pervasive that it has become almost impossible to think outside of it. You are either a “traditional” medium, in which case you are presumed to be dying with the dignity of a regional newspaper, or you are a “digital” medium, in which case you are presumed to be the future and are granted the benefit of every doubt, including the doubt about whether anyone is seeing your ads.

I think the binary is wrong. I think it has been wrong for a while, and I think AI is about to make its wrongness impossible to ignore.

The real divide is not analog versus digital. It is trusted versus mistrusted.

This is not a philosophical distinction. It is not the kind of thing you say on a panel and then everyone nods thoughtfully and then nothing changes. It is an economic distinction with research behind it. The Institute of Practitioners in Advertising and the Financial Times ran a study that found sixty percent of consumers are more likely to trust a brand that advertises in a trusted media environment. Sixty percent. The mechanism is so straightforward it almost sounds naive: if you trust the medium, you extend that trust to the message. The researchers call it the halo effect. It is not new. It has been studied for decades. What is new is the degree to which the untrusted end of the spectrum has deteriorated. The halo effect was always an advantage. Now it is becoming a moat.

Consider what makes a medium trusted. It is gatekept. Its content is curated and professionally produced, not algorithmically generated or user-submitted or created by a bot pretending to be a person pretending to have opinions about kitchen appliances. It exists in a verifiable context. You can see it. You can point at it. You can confirm, with your own eyes, that it is real and that it is there.

A billboard cannot be deepfaked. This is such a simple observation that it feels almost stupid to write down. But in an era when the Prime Minister of Israel can be convincingly simulated in multiple videos that your uncle shares on Facebook claiming he’s dead without a second thought, the inability to be faked is a competitive advantage of historic proportions.

A subway wrap does not have a bot problem. A transit station domination is not competing with AI slop for your attention. When you and I drive down the same highway, we see the same board. That sentence is so simple it sounds meaningless, but in 2026, it might be the most radical value proposition in all of advertising. Because it means the impression is real. The audience is real. The environment is real. And in a media landscape where the word “real” has become genuinely difficult to apply to most digital inventory, that is not nothing. That is, in fact, everything.

Out-of-home advertising is a public medium. It is universal. It is unfiltered. It cannot be divided by algorithms or personalized into the filter bubbles that have made so much of the digital ecosystem feel like a series of increasingly narrow hallways, each one wallpapered with content that was generated specifically to confirm what you already believe and sell you something you probably don’t need. OOH is embedded in the physical culture of cities, in the daily routines of millions of commuters, in the architecture of actual places where actual people actually go. It is, in the most literal possible sense, real.

A Basis Technologies roundtable published in December 2025 brought together three industry veterans who arrived at the same conclusion from different directions: trust has now emerged as equally important to price and quality in consumer purchase decisions. Equally important. This should change media allocation. Whether it will is a different question, and one that depends on whether the people making the allocations are willing to question the binary they have been operating under for the last fifteen years.

III. Your AI Agent Will Buy You Toothpaste. It Will Not Build You a Brand.

Now here is where it gets interesting. Because the trust argument is not just about the present. It is about a future that is arriving considerably faster than most media plans account for, and which threatens to make the present look quaint.

Agentic AI - the kind where software doesn’t just assist your decisions but makes them on your behalf - is no longer a conference keynote abstraction. It is a product category. Visa and Mastercard are building payment infrastructure for it. OpenAI and Google are racing to establish competing protocols. Consumers are beginning to delegate shopping research, product comparison, and eventually purchasing to AI agents that act autonomously on their behalf. “Buy me the best blender under $200.” “Rebook my flight with a change fee under $150.” These are not hypothetical prompts. They are current product features.

Thirty-eight percent of consumers already use AI when shopping. Eighty percent expect to use it more. Those are IAB numbers. They are from this year.

But here is the part that matters, the part that should be underlined in every brand strategy document in every CMO’s office: only twenty-four percent of consumers say they are comfortable letting AI actually complete a purchase. That’s from Bain & Company. The gap between “I’ll use AI to help me decide” and “I’ll let AI decide for me” is enormous. And the variable that determines which side of the gap a consumer falls on is trust. Trust in the agent, yes. But more fundamentally, trust in the brand.

Adobe’s 2026 Digital Trends report uncovered a perception gap so wide it borders on institutional delusion. Forty-nine percent of organizations believe their customers will eventually want AI agents to be their primary way of interacting with brands. Nineteen percent of customers agree. Thirty points of daylight between what companies think consumers want and what consumers actually want. A third of consumers say they would disengage entirely upon discovering that content is AI-generated. Thirty-seven percent say the same if they discover they are talking to AI when they expected a person. These are not anti-technology zealots. These are normal people expressing a normal preference for knowing what is real.

Here is what I think happens next, and I want to be precise about this because I think it matters.

As AI agents take over more of the consideration and conversion funnel - the research, the comparison, and the deal-finding (the boring middle section of the purchase journey that nobody has ever enjoyed), brand building becomes simultaneously harder and more important. It becomes harder because you cannot build emotional salience through an algorithm. You cannot create the kind of deep, intuitive trust that drives brand loyalty through a chatbot interaction that the consumer may or may not realize is happening. The warm feeling you get when you see a brand you love on the side of a building in your neighborhood - that is not replicable in a programmatic display unit.

But it becomes more important because when your AI agent asks you what to buy, you will tell it the name of the brand you trust. You will say “get me the Nike ones” or “book the Marriott” or “I only want Apple.” And you will trust those brands not because their agents are better, but because you have encountered them in real life, in contexts that felt trustworthy, over a period of time long enough to form a genuine opinion. The brands that have planted themselves in your memory through physical, cultural, unavoidable presence will be the ones that survive the agentic filter. The rest will be left hoping that an algorithm recommends them, which is to say they will be hoping for the best, which is to say they will be in trouble.

Harvard Business Review published a piece in March of this year called “Preparing Your Brand for Agentic AI.” The thesis is that brands must adapt to a world where consumers increasingly rely on AI for product discovery and recommendations. I agree with the thesis. I would add one thing: the best way to prepare for a world where machines mediate the brand-consumer relationship is to build the relationship in the one place machines cannot mediate. In real life. On a wall. On a train. On a board that you drive past every morning and that, without your conscious permission, has become part of the geography of your life.

IV. Virtual Companies Are Spending Real Money to Prove They Exist

There is a tendency, when people in out-of-home talk about AI, to talk about it exclusively as a disruption. A threat. An incoming asteroid. Something happening to the industry, requiring defensive positioning and careful scenario planning and a lot of nervous conferences in hotel ballrooms.

This is only half the story, and it is the less interesting half.

AI companies are becoming one of the most significant advertiser categories in the medium. Not theoretically. Not eventually. Right now, today, in a way you can see with your own eyes if you step outside and look up.

Walk through any major airport in the United States. Drive through San Francisco, which has become so saturated with AI advertising that it looks like the physical manifestation of a venture capital pitch deck. Ride the New York City subway. The boards are covered. Not just the ones you have heard of - the Anthropics and the ChatGPTs and the Googles with their Gemini campaigns - but B2B healthcare AI startups, vertical SaaS disruptors, wearables companies, cybersecurity firms, legacy enterprise software companies scrambling to communicate that they have integrated AI into products that were built before most of their current employees were born. Every stage of the AI lifecycle is represented: pre-revenue companies spending for fundraising visibility, growth-stage firms acquiring their first customers, established players defending market position against challengers who did not exist eighteen months ago. The category is growing fast enough that major OOH companies have created dedicated industry verticals, with heads of industry and specialized sales teams, just to manage the demand.

This makes intuitive sense if you think about it for more than a few seconds. AI companies are, almost by definition, virtual businesses. They exist as code, as interfaces, as subscription tiers, as conversations you have with something that may or may not be a large language model. They do not have storefronts. Many of them do not have physical offices that anyone outside the company has ever visited or could identify in a photograph. They are, in the most fundamental sense, not real in the way that consumers experience reality. And they know this. Which is why they are spending aggressively to become real. To be seen. To exist in the physical world, on a surface you can touch, in a city you can name, next to a highway you drive every day.

The out-of-home medium is giving AI companies the one thing they cannot generate for themselves: a physical presence. A body. An existence in the world that does not require a login. And in doing so, it is demonstrating exactly the value proposition I have been describing. The medium is trusted because it is real. The advertisers are flocking to it because they need to be real. The logic is circular, and it is correct.

V. The Industry That Forgot to Sell Itself

I should be honest about something. The out-of-home industry has not exactly covered itself in glory over the last decade.

U.S. OOH ad spend has declined from roughly four percent of total media spend to about two and a half percent. To put that in context, China is between six and seven percent. Germany is nine percent. Australia has gone from four to six percent. The United Kingdom is increasing. We are going the other direction, which is not the direction you want to be going in a country that is the largest advertising market in the world by a factor that makes the comparison almost embarrassing. This did not happen because the medium stopped working. Every piece of evidence says it works better than it ever has. It happened because the industry failed to keep pace with how media is bought, sold, and measured. It happened because when every other medium was modernizing its infrastructure, integrating with programmatic platforms, investing in attribution and measurement, and generally behaving as though the future was coming, out-of-home was still, in too many cases, operating like it was 2005, asking buyers to make phone calls and review PDFs and take the medium’s effectiveness on faith.

The buyers stopped having faith. You cannot blame them.

Seventy-five percent of U.S. ad spend is now digital. More than eighty percent of that is traded programmatically. Out-of-home is under twenty percent programmatic. Think about what that means operationally. If you are a media buyer sitting at a trading desk, planning an omnichannel campaign, and the out-of-home button is not even on your screen alongside CTV, streaming, and retail media, then OOH does not exist for you. It is not that you have evaluated it and decided against it. It is not that you have weighed the CPMs and the attribution data and concluded that some other channel offers better value. You have never had the opportunity to evaluate it at all. You have never seen the button. That is a distribution problem, not a demand problem. And distribution problems, unlike demand problems, are fixable.

It is changing. Programmatic DOOH ad spend is projected to reach $1.35 billion by 2026, growing over twenty-two percent year-over-year. The industry is replacing its legacy audience measurement system, which had fallen behind, with modern alternatives backed by research firms whose names enterprise marketers actually recognize and trust. The OAAA and Winterberry Group published research just weeks ago showing that ninety-eight percent of marketers now view OOH as a core or supporting component of their connected commerce strategies. Eighty-six percent plan to increase their OOH investment over the next two years. These are not aspirational numbers from a trade association trying to justify its budget. These are directional indicators of where money is about to move.

And there is the value argument, which is so lopsided it almost feels unfair to mention in the context of the thirty-six-cents-on-the-dollar conversation. OOH has the lowest CPMs of any advertising medium. Digital out-of-home drives 3.2 times more neurological response and memory activity than traditional static formats. That’s from a neuroscience study, not a survey. Eighty percent of consumers say they are likely to take action after seeing a creative, visually engaging OOH ad. Nearly half search for the advertiser. Almost a quarter make a purchase.

These are measured outcomes from real campaigns seen by real people. People with bodies, in cities, doing things. Not bots. Not synthetic impressions. Not thirty-six cents on the dollar. The whole dollar. Delivered to an actual human being who might actually buy something.

The gap between what OOH delivers and what the industry charges for it is the single greatest arbitrage opportunity in modern advertising. It will not last forever.

VI. The Medium Is Ready. The Question Is Whether Anyone Has the Nerve to Say So.

I have spent most of this article arguing that the future of media is not about analog versus digital. It is about trusted versus mistrusted. I believe this. I believe the data supports it in ways that would be dispositive in any other industry but that in advertising, which has a unique capacity for ignoring inconvenient evidence, remain somehow up for debate. I believe the next five years will prove it out in ways that are obvious in hindsight and surprising to no one who was paying attention.

But here is the part that keeps me up at night. The out-of-home industry has the best hand it has been dealt in a generation, and it could still fold.

It could fold by being timid. By continuing to accept low-single-digit growth as normal, as though there is some law of physics that says outdoor advertising is structurally constrained to grow at the rate of GDP while every platform built in a dorm room gets to double every eighteen months. By failing to invest in the measurement infrastructure that enterprise marketers require, not because they are being unreasonable but because they have been burned too many times by media that could not prove what it claimed. By treating programmatic integration as a nice-to-have feature on next year’s roadmap instead of what it actually is, which is an existential necessity. By showing up at every conference talking about “traditional out-of-home” in a tone that suggests mild embarrassment, instead of talking about trusted brand building in real life in a tone that suggests absolute conviction.

Philip Kotler, who has probably been quoted in more marketing textbooks than any other living person and who is by now presumably tired of being quoted, once said something that I think about often. You are either a brand or a commodity. There is no middle ground. If you want to be a brand - if you believe in the value of trust, of premium positioning, of a story that people remember and choose to believe in and eventually tell to their AI agent when it asks what to buy - then you need to find the media that are trusted. You need to show up where trust lives.

In a world of synthetic content, algorithmic curation, and AI-mediated commerce, that is not going to be a website that is fifty-one percent bots. It is not going to be a programmatic supply chain that loses sixty-four cents of every dollar before the ad even loads. It is going to be a billboard on the highway, a screen in the subway, a wall in the city where you live. It is going to be real life. It has always been real life. We just forgot, for about fifteen years, to say so.

The medium is ready. The question is whether the industry is brave enough to say so out loud, to the people who control the budgets, in a voice that does not apologize for what it is.

I think it is. But then again, I would.

A tip of the hat to Nick Brien, CEO of OUTFRONT Media, whose remarks at the Morgan Stanley Technology, Media & Telecom Conference in March 2026 planted the seed for much of the thinking in this piece. His framing of the future as “trusted environments versus mistrusted ones” and his insistence that the OOH industry stop apologizing for what it is and start going on offense struck a nerve. This article is an attempt to follow that nerve to its logical conclusion.