Made by Machines. Meant for Humans. Winning Trust in the Cre(ai)tive Economy
AI agent traffic grew nearly 8,000% year-over-year. More than half of the internet is now synthetic. Those aren't projections, they are findings from HUMAN Security's latest annual report, 2026 State of AI Traffic & Cyberthreat Benchmark Report, and they set the stakes for a conversation we hosted alongside AgenticAdvertising.org (AAO): if machines are now making, moving, and buying much of what consumers see, how do we make sure it's still meant for humans?
Our panel brought together five leaders sitting at every point of the value chain: Sarah Acker, SVP Marketing at HUMAN, Mike "Chevy" Chevallier, SVP of Advanced Audio and Data Revenue at iHeartMedia; Priti Ohri, CEO and co-founder of Advertible; Christina Cubeta, CMO of Scope3 and founding adviser to AAO; and Amanda DeVito, CMO of Butler/Till. What emerged wasn't a debate about whether the agentic era is coming. It's here. The real question is whether we build it on trust.
The train has left the station — and nobody knows the destination
Every panelist agreed on the diagnosis: agents have been deployed across the industry at explosive speed, and largely in a non-standardized way. As Chevallier put it, the conversations happening now are about how to reign the horse back in — establishing standards and protocols before the wild west becomes permanent. Priti drew the parallel to the early days of programmatic: enormous energy, enormous disruption, and a lot of people still figuring out what to actually do with it.
DeVito made the observation: "The train has left the station... no one actually knows where the destination is." That uncertainty is real, and so is the fear inside agencies and brands. But the panel converged on a shared conviction: the risk of inaction now outweighs the risk of experimentation.
This is exactly the moment AAO's whitepaper, Building the Future of Marketing, addresses. It describes the shift from the programmatic to agentic as an architectural change, not an incremental one, and it defines the Cre(ai)tive Economy as the point where brand strategy, creative, and media stop being separate silos and start operating as one continuous, optimizing loop. In the report's framing, the marketer's job moves from authorship to orchestration: less time executing tasks, more time designing systems, setting constraints, and governing the agents that do the work.
Trust isn't bolted on. It's built in.
Cubeta made a point that should reframe how the industry thinks about this transition: the temptation is to bolt agents onto the current system — the same system in which marketers lose roughly 60 cents of every dollar to inefficiency and opacity. That's not the promise of the agentic era. The promise is a rebuild from the ground up, which is why AAO launched the Ad Context Protocol (AdCP).
Priti described what that rebuild looks like in practice: checkpoints between the brand, the buy agent, the sell agent, and the creative agent, so that agents operate only on approved assets and within declared guardrails. The whitepaper details the machinery behind this — standards that let brands establish verified identity and declare what their agents are and aren't permitted to do, in a machine-readable format. Misuse becomes visible and auditable. Humans are embedded architecturally, not as an afterthought.
But Cubeta pushed the trust conversation past infrastructure and into something more uncomfortable for brands: control is not the same thing as intentionality. Brands say they fear losing control to AI. What they should fear is scaling a brand they haven't defined. If you don't know your voice, your audience, and how people actually internalize your brand, AI won't wreck it — it will simply amplify whatever is (or isn't) there, at a speed that used to be unimaginable. Garbage in, garbage out, at machine scale.
Guaranteed human, powered by agents
No one on the panel embodied the "made by machines, meant for humans" tension better than iHeartMedia. On the consumer side, their promise is literally "guaranteed human" — real voices, real hosts, no AI-generated slop, because their research shows roughly 90% of consumers want exactly that. On the operations side, through their ad-tech arm Triton (a founding contributor to AdCP), they are fully leaned into agentic systems to modernize planning, selling, and measurement.
That dichotomy isn't hypocrisy, it's the model. Machines handle the coordination; humans own the relationship, the judgment, and the accountability. As Chevallier framed it, whatever gets deployed on the automation side carries human governance behind it, because the consumer promise demands it.
The same logic applies to disclosure. The panel noted that consumers don't hate AI content as much as they hate ambiguity: the cognitive burden of wondering whether something is real. Transparency removes that dissonance. Tell people something was made or assisted by AI, and most will lean in. Make them guess, and you invite backlash.
The human stays in the loop — and levels up
Where does that leave the people? Every panelist landed in the same place: humans don't leave the loop, they move up in it. Task-level work gets absorbed by agents; humans elevate to strategy, vision, and system design — precisely the orchestration shift the AAO report describes, along with the education and certification infrastructure needed to professionalize it.
Cubeta offered the panel's best defense of human creativity: give three AIs the same dataset and the same question, and over time they'll converge on the same answer. Brief three film directors and you'll get three different scripts. Humans hunt for the thing that's never been done. That instinct isn't threatened by the Cre(ai)tive Economy, it's the scarce input the whole system depends on.
And DeVito added the cultural dimension leaders can't ignore: this transition creates real inequity and anxiety inside organizations. The answer isn't false confidence. It's transparency — "confidence is knowing what you don't know" — plus the psychological safety for people to challenge briefs, flag bias, and experiment.
What winning looks like
Asked to describe success five years out, the panel painted a consistent picture: a force multiplier that removes the mundane and re-engages talent; simplicity that gives practitioners space to work from craft rather than crisis; democratized access to audiences and inventory that finally delivers the right message to the right person at the right moment; and a rebirth of the open internet, where premium content surfaces thrive because agents can discover and value them fairly.
That future doesn't arrive on its own. It requires shared standards so brand marketers aren't managing a chaos of incompatible protocols. It requires provenance and interoperability. To know who (and what) you're transacting with, and trusting that the transaction is authentic. And it requires an industry willing to treat trust not as a compliance checkbox but as the core product.
The Cre(ai)tive Economy will be made by machines. Whether it's meant for humans — whether it earns their attention, their confidence, and their trust — is a choice we're making right now, together.
To re-live this panel at Cannes Lions, see the full recording here. To go deeper on the Cre(ai)tive Economy and the agentic transformation of marketing, read AAO's full whitepaper, Building the Future of Marketing. To learn more about AgenticAdvertising.org and to join the community building the future of AI and advertising, see our membership program here.