AI Ads are Blurring the Line Between Assistance and Influence
As AI shifts digital interfaces from static pages to live conversations, Nic Baird and Kevin Baragona reveal that advertisers and publishers who align ads with real-time user intent will define the next revenue engine of the internet.
Topics
For years, digital advertising operated within a predictable architecture.
A page existed. Inventory was defined. Ads were placed, optimised, and measured within that fixed frame. It was a system built on certainty, that is, one that rewarded scale, precision, and control.
That system also scaled into a $600+ billion global industry, with programmatic advertising alone accounting for nearly 90% of digital display spend. Its success came from standardisation — repeatable formats, measurable surfaces, and well-understood user behaviour.
But AI has quietly removed the page.
What replaces it is not another format, but a fundamentally different environment: a conversation. Fluid, intent-rich, and constantly evolving. Here, users do not browse; they ask, refine, and explore. And in doing so, they expose something advertisers have always chased: Real-Time Intent.
The shift is subtle, but its implications are not. As the interface changes, the infrastructure behind it is forced to follow. And increasingly, the old playbook — built for static surfaces — is showing its limits.
When Infrastructure Meets Reality: Why Legacy Ad Models Break
At its core, legacy ad infrastructure was never designed for conversation. It was designed for placement.
“There’s a page, the page has slots, you fill the slots,” says Nic Baird, Co-Founder, Koah. “In a live conversation, that surface doesn’t exist. So everything breaks at once.”
The breakdown is not gradual. It is systemic.
- Targeting loses relevance:
Legacy systems depend on historical signals: Past Searches, Behavioural Patterns.
But in conversational AI, intent is immediate and explicit. “The user is telling you exactly what they need right now,” Baird notes. And yet, traditional systems have no mechanism to truly interpret that signal in real time.
- Placement becomes fragile:
In static environments, poor placement is a missed opportunity. In conversations, it is a disruption. “A forced insertion at the wrong moment… destroys the trust that makes the app valuable,” he explains.
- The funnel no longer fits:
Search ads assume purchase intent. Display assumes awareness. But AI conversations sit somewhere in between — exploratory, iterative, unresolved. “Legacy systems weren’t built for that moment because that moment didn’t exist as an ad surface before.”
The conclusion is difficult to avoid. This is not a format problem. It is an architectural one. And as Baird puts it plainly: “You can’t retrofit a static-page logic into a live conversation. It has to be native by design.”
The Trust Equation: Transparency as a Performance Lever
If structure is breaking, trust becomes the stabilising force.
The concern among publishers is clear: Ads embedded within AI responses risk blurring the line between assistance and influence. But the tension, Baird argues, is often overstated — a false tradeoff shaped by past mistakes.
“The publishers who’ve burned trust with ads did it through unmarked placements, irrelevant interruptions, creative that looked like part of the response but wasn’t.”
In other words, the issue is not advertising itself, but poor design.
When those mistakes are corrected, the dynamic shifts. Ads that are clearly labelled, contextually aligned, and introduced at the right moment do not feel intrusive. They feel useful. Even welcome.
“What we’ve found is that when an ad is clearly labelled, contextually relevant, and shows up at the right moment… users don’t resent it, they actually engage with it.”
The data reflects this shift. Engagement rates in conversational environments are outperforming traditional formats, not despite transparency, but because of it.
The implication is subtle but important. In a conversational interface, relevance is not just a targeting advantage. It is a trust mechanism. And trust, in turn, becomes a performance driver.
From Banners to Belonging: Why Native Ads Feel Like an Upgrade
For some platforms, the move toward LLM-native advertising has not been an experiment, but a necessity.
Kevin Baragona, Founder, DeepAI, recalls a reality many digital businesses recognise but rarely admit: banner ads were already failing. “It was basically impossible to be considered premium with traditional banner ads,” he says.
The users who stayed were not engaged; they were indifferent.
The real tension was not whether to introduce ads, but whether to continue with a format that actively degraded the experience.
- From interruption to integration:
Banner ads existed around the experience. Native ads exist within it. “The ad surfaces in a moment of natural interaction,” Baragona explains, rather than sitting on the periphery of attention.
- From visibility to usefulness:
The shift is not about hiding ads, but making them relevant. Contextual recommendations feel less like promotions and more like assistance, aligned with what the user is already trying to do.
- From tolerance to acceptance:
“The question won’t be ‘ads or no ads,’” he says. “It’ll be ‘good ads or bad ads.’” When ads add value, they stop being a compromise and start becoming part of the product.
Interestingly, earlier concerns about disruption proved overstated. Users had already learned to ignore traditional ads entirely. Native formats, by contrast, engage precisely because they appear at moments of intent.
What emerges is not a cleaner version of advertising, but a more integrated one, where ads do not compete with the experience, but contribute to it.
The New Backbone of the Internet Economy
As AI continues to reshape digital experiences, advertising is not disappearing. It is relocating and redefining itself in the process.
“There’s very little friction between native ads and a premium experience when the ads are genuinely contextual and useful,” Baragona observes. The implication is clear: the future of monetisation will not be determined by whether ads exist, but by how well they align with user intent.
Baird’s approach points to a similar shift at the infrastructure level. By relying on real-time conversational context instead of persistent user profiles, advertising becomes both more relevant and less invasive.
“You don’t need their browsing history from three months ago if someone is telling you exactly what they need right now,” he says.
This convergence, relevance without surveillance, monetisation without disruption, is what makes LLM-native advertising structurally different.
And structurally durable. As Baragona puts it, “ads are a structurally important part of the economy.” As AI interfaces become the front door to the internet, that economic layer will follow. Not as an afterthought, but as a foundation.
The question now is not whether this model will scale. It is who will define what “good” looks like when it does.
ALSO READ: When Brands Lead with Purpose – Connection Through Community-Led Campaigns
