As Cannes Lions nears, the conversations happening in the run-up to the media and entertainment industry’s most anticipated event tell you something about where the industry’s head is. Unsurprisingly, talk is less about individual campaign wins and more about whether the teams behind those campaigns are making the structural decisions that will determine whether they are positioned for lasting long-term advantage.
As AI continues to compress the traditional agency model, it is equally opening up significant ground for the agencies willing to move toward higher value – accelerating a divide already forming between those that have made a decisive strategic choice to capitalise on that opportunity, and those still treating AI adoption as a capability to acquire rather than a structure to build around. We speak to agencies every day, and they’re telling us just how acutely they’re feeling this pressure – and how strong the appetite is to respond to it differently.
It is a tension that Cannes, now as much a technology conversation as a creative celebration, is perfectly placed to surface. Ahead of this year’s show, one question requires a categorical answer.
What are the agencies that are actually getting this right doing differently?
The traditional agency model is feeling the pressure from all sides
In 2025, Gartner’s CMO Spend Survey found that nearly four in ten CMOs were already planning to cut agency budgets as a direct result of AI-driven productivity gains, and more tellingly, over a fifth reported that generative AI had specifically reduced their reliance on external agencies for creativity and strategy. And this year, the picture has sharpened further as JP Morgan’s recent analysis confirmed what the industry has already been feeling. Brand pressure for measurable ROI and competition from self-service platforms is forcing agencies into build-or-buy decisions about capabilities, with a clear divide emerging between what platforms can now automate and what still requires human strategic oversight.
But the capability squeeze is only part of it. The data imperative is pushing brands toward more direct relationships with audiences, creators, and their own customers. Every layer of intermediation creates exposure, with customer relationships held at a distance, data and IP passing through third parties. And as the creator economy continues to restructure, that shift doesn’t naturally route through a traditional agency model either. Brands moving upstream with creators and building persistent audience relationships are increasingly doing so directly and effectively.
This is what the pressure looks like from the outside, but the more interesting question is how agencies are responding to it, because this is by no means the whole story. The same forces compressing the traditional model are creating genuine space for agencies willing to move toward higher value work, deeper strategic relationships, and the kind of creative and commercial oversight that platforms and automation simply cannot replace. But this is only achievable through the kind of structural reorganisation around AI that enterprises are often struggling to make.
Why incremental AI adoption is losing ground
The temptation for most agencies has been to reach for the safe option. To use AI to make existing processes faster, cheaper, and marginally more efficient, and label it transformation. The problem with this approach is that optimising what exists and reorganising around what’s possible are not the same thing, and the gap between them is where competitive advantage is now being won and lost.
Truly AI-native competitors aren’t winning by using more AI. They’re winning by treating autonomy, data flows, and decision latency as structural advantages, built in from the ground up. Which means agencies still bolting capability onto legacy structures aren’t just moving slowly, they’re now moving in the wrong direction. As boards and CEOs push for sharper choices and demonstrable returns, experimentation is no longer considered a strategy. Strategy today hinges on having the right foundations in place to hold the ambition itself.
The agencies coming out ahead have understood this. They are asking how their existing model needs to change in order to stay ahead of what clients will demand next, and then executing on that change strategically.
The agencies leading the AI opportunity
Instead of defending existing territory, these agencies are reorienting their entire business model around AI in order to capture the ground that’s opening up.
Monks is a clear example of what implementing a truly effective AI strategy can look like in practice. Rather than layering AI onto existing workflows, they rebuilt around it, launching monks.flow to deploy AI agents across the full marketing workflow and, in doing so, shifting team resources away from execution and toward the strategic direction, creative judgement, and commercial oversight that actually drives measurable client outcomes.
This kind of structural commitment also sends a signal – because too often, agencies are advising transformation while leaving their own operations untouched. Alignment between what an agency tells its clients and how it actually runs will increasingly be the credibility test clients apply.
What I expect Cannes to surface this year
As we head to Cannes, I expect to see more of this realisation taking hold across the agency landscape, and more broadly across the industry in the months that follow.
Needless to say, the baseline expectation has already shifted – arriving with an impressive AI demo no longer moves the needle. And as clients and peers begin to assess the structure underneath, whether existing operations have been genuinely interrogated, and whether commercial strategy has been reformulated around today’s possibilities, will be the measures that matter.
More than that, I expect Cannes to surface some of the sharpest thinking yet on where human value commercially sits inside these new operating models – bringing strategic counsel, creative judgement, and cultural fluency to the fore as the competitive differentiators that agencies who understand their clients are building around.
Building the foundations ambitious AI strategy demands
Translating this ambition into commercial reality must become an infrastructure question as much as a strategic one. The agencies that get this right won’t be stitching together disconnected tools or managing the complexity of a fragmented AI stack built on yesterday’s infrastructure. Instead, they will need to build, or partner with the organisations that can provide the foundational systems that enable them to move at the speed the market now demands.
It is why we built Stelia AI OS; the full-stack operating system for production-grade AI, giving agencies and their clients the speed to move fast, the scalability to drive real revenue from existing data, and the adaptability to capitalise on what’s next without being constrained by the platforms of yesterday. A governed, sovereign, scalable platform designed to deliver the foundations required to operationalise AI effectively.
I’m heading to Cannes Lions expecting the most consequential version of these conversations yet. The event has always reflected where the industry’s ambition sits, and this year, executing on that ambition has real urgency behind it.
