AI won’t – Speaking at Cannes Lions, Amy Lanzi—CEO of Digitas North America—argued that the advertising industry’s AI hype echoes earlier promises in programmatic. She said the real work is building marketing systems, using data and analytics to support creators, and res
By the time Cannes Lions starts to feel like a familiar loop, the room has usually already decided what it wants to believe. This year, it wanted AI.
On stage at the Uber Villa in the South of France. Amy Lanzi—CEO of Digitas North America. part of Publicis Groupe—didn’t dress the message up. She described the pitch-season energy she’s seeing around “free AI” and other blockbuster promises as a kind of chaos that doesn’t help anyone—because advertising is still “a people business. ” and deals built on hype turn into “a people problem” long-term.
The setting couldn’t be more on-brand for the moment: Cannes brings the advertising industry together every year. and this time the conversation was dominated by AI promises made to brands. Lanzi pointed to a Publicis video, “The Wrong Promises,” described as a “fake ad” documentary-style piece. It presents vignettes of claims people make in pitches. including “you don’t have to pay us until you win a Lion. ” and carries the warning “this is real” at the bottom.
Then she returned to the pattern she thinks the industry can’t stop repeating. When people talk about AI. she said. she goes back to programmatic—the era when automation was supposed to make everything work “magically. ” before reality forced the industry to admit it still “needs people. ” nuance. and brand-and-market understanding.
In her view, the AI story is just the new version of that promise: automation offered as a solution, with partners—agency partners and tech/platform partners—selling wild outcomes that may be delivered only on paper.
Publicis and Digitas, she said, aren’t pretending they’re standing still. Digitas, she explained, has invested in AI—“two years” in—and has also bought Influential in 2024, a move aimed at analyzing how the creator ecosystem is doing.
But the way she describes those investments is less about replacing humans and more about changing workflows.
Digitas AI. Lanzi said. started with an effort to “make our people better. ” turning daily work into opportunities for agents—imagining what employees could build to take on tasks so staff could focus on higher-stakes creative and strategy. She tied it to an ongoing belief about hacking mentality in the industry: young talent. curious and building toward real working problems that later become business problems.
Structure is part of the pitch, too. Asked whether AI change comes top-down or bottom-up, Lanzi pointed to changes Digitas has made to its roles and organization. She said Digitas has a chief intelligence officer, a chief systems officer, and a chief transformation officer.
She framed those roles around the idea that agencies need to scale faster in the parts underneath their services—leveraging AI as “intelligence” across an intelligent platform that supports agents and is used by employees. The systems officer shift, she said, reflects that clients increasingly talk in terms of “marketing systems,” not standalone campaigns.
And the transformation officer. she said. is meant for clients transforming CMOs into CGOs—moving marketing leadership from a role once focused on making “marketing magic” through campaigns and media investment into something more capability-driven. In her telling, that’s both a tech problem and a people problem.
Her critique of the hype doesn’t stay in the abstract. She pushed directly back on one of the most popular platform narratives in the industry: the idea that “creative is the new targeting.” She didn’t hide her discomfort with the phrase.
Lanzi described “creative” as emotional and “lovely and magical. ” and said targeting is something that doesn’t belong to the same sentence. In her view. what brands that perform well are doing instead is using context—especially their understanding of Gen Z—to deliver creative that fits the moment without assuming there is unlimited volume of content and impressions.
“The brands that are doing really well. ” she said. “are deciding not to do that or to do it in a way that feels so contextually relevant.” In other words: not more content. but learning. She connected the dots back to measurement. arguing that brands need to keep asking what worked—whether it gained attention from the specific “growth audience” they’re trying to reach.
Where the platform hype gets most dangerous, she said, is when it becomes an argument that brands should simply hand creative and targeting to platforms.
She illustrated the idea with her own blunt example: a scenario where a brand provides an image of a shampoo bottle and a platform automatically places it in the right setting—she jokingly imagined a truck-explosion scene and shampoo appearing at the end.
Then she returned to how real advertising can’t just be “left to go on its own.” People are complicated. she said. If you see the same ad too many times you don’t just lose efficiency—you lose goodwill. Without someone actively suppressing messaging and understanding consumers, the result can be more than waste.
That’s where partnerships, she argued, still matter. She described a strategy angle that comes up again and again in modern marketing: there are many players and they shift in and out depending on where audiences are spending their time and how they intersect with a category.
Digitas’s purchases and alliances are meant to help with that reality. She pointed to Publicis buying LiveRamp—described as a “giant data platform”—and acknowledged that the deal wasn’t done. Her point wasn’t that platform automation is useless; it was that data. identity. and cross-platform measurement are still where agencies fight for leverage.
The question hanging over all of it is what comes next when search and discovery get compressed by AI.
Lanzi said she’s already watching changes on the publisher side—traffic from Google going down—and she described a similar concern around discovery traffic, saying publishers are looking for ways to create custom content less dependent on search.
Her prescription was simple: if a publisher has an authentic, true voice, people will follow and come find them. Trying to “fight it out in search,” she said, is harder.
She also said brands are increasingly revising websites—sometimes for a purpose that overlaps with generative engine optimization (GEO). which she described as the idea of being “known” rather than just “found.” In her description. websites are shifting from random information hubs into “knowledge sites. ” meant to feed LLMs while also building brand legitimacy and giving consumers something real.
The stakes are bigger than marketing dashboards. She argued that having owned assets matters as commerce shifts. Amazon. she said. is the “Google in this story” for many packaged goods brands. and she noted how marketplace dominance can make brands vulnerable to commoditization even when those brands have invested in a category.
Her comments also kept returning to creators—arguing that the creator economy isn’t going away, even if some of the pricing and dynamics are changing.
At Cannes, she said, she expected creator rates to fall as AI-generated content grew and platforms became more punishing about what works in algorithms. Instead, she said she’s seeing “the cost per minute of video” rise again.
In her telling, supply and demand explain it, and fame matters. She also tied the creator trend to a broader platform-driven shift: creators see themselves as marketers, and top creators sometimes carry their own products.
That’s the opportunity—and the tension.
Lanzi described scaled creators as businesses that can need agency help. She brought up examples of creators turning into enterprise brands and said that when that happens. they start asking the kinds of questions traditional marketers ask: how to invest media. how to grow in unowned ecosystems. how to handle distribution when shipping physical products gets expensive.
She named an “influencer cliff” thesis associated with Carmen Vicente. saying the idea is that branded content works while audiences tolerate ads embedded in content creation. But when creators pivot into direct selling and brand integration gets too heavy. audiences can feel the misalignment—and creators end up backing away. issuing apologies. and “falling off the cliff.”.
To understand who can cross into selling their own products, she said a unique product is important—not a “me-too” offering—and it helps when the product feels like a natural extension of who the creator is.
Even her comments about GEO landed back on human stakes: Lanzi doesn’t believe AI creative will become good enough to replace the complex storytelling that humans produce.
In her view, AI can speed concepts up and make higher production more efficient—but “when I think about big creative and storytelling,” she said, “it still needs humans.”
At the end of the conversation, Digitas’s focus sounded less like an AI roadmap and more like a systems mission: marketing systems that tie together media needs, creative needs, and CRM, so brands can build growth engines.
It was a message delivered in the middle of one of the industry’s most hype-heavy weeks. And for Lanzi, the lesson wasn’t to ignore AI. It was to stop treating it as a shortcut out of the mess.
AI may help people work faster. But the chaos in advertising, she left little doubt, is not something a new tool can automate away.
Digitas Amy Lanzi Cannes Lions Publicis AI advertising creator economy marketing systems programmatic GEO LiveRamp Influential TikTok shop marketing analytics