- Tension: The advertising industry’s most prestigious festival awarded a Grand Prix to a campaign that independent auditors found had never actually run in the way it claimed.
- Noise: Coverage has framed this as an AI problem — new tools, new detection required — obscuring the older issue: the awards system had no reliable mechanism to verify that honored work ever touched a real consumer.
- The Direct Message: An industry that cannot verify its own best work has a credibility problem that predates AI and will survive any rule change that doesn’t address the incentive to fabricate.
To learn more about our editorial approach, explore The Direct Message methodology.
Sometime before the Cannes Lions submission deadline in 2025, someone at the Brazilian agency DM9 made a decision. The campaign they were entering — “Efficient Way to Pay,” for the Consul brand under Whirlpool — included footage from a CNN Brasil broadcast that had been manipulated with artificial intelligence. The campaign had not run in market in the form the case study presented. After independent auditors found the work was not what it claimed to be, DM9 withdrew the entry. Cannes Lions revoked all twelve awards the agency’s three campaigns had accumulated: one Grand Prix, three golds, four silvers, and three bronzes.
Twelve awards revoked from a single agency at a single festival — a scale of withdrawal the event had not seen before, at least in the public record.
What was submitted
“Efficient Way to Pay” won the Creative Data Lions Grand Prix — one of the more technically scrutinized categories at the festival, where the evidentiary bar for demonstrating data-driven creative effectiveness is supposed to be high. Alongside it, DM9 entered “Plastic Blood” for OKA Biotech and “Gold = Death” for the Urihi Yanomami Indigenous health initiative. All three were found to contain problems with evidentiary integrity. The AI manipulation in the Consul campaign was the clearest case: footage from a major Brazilian news broadcast, altered to construct evidence of campaign impact that hadn’t occurred.
What was fabricated, or significantly misrepresented, was not the underlying creative work itself but the proof of its effect — the case study materials designed to demonstrate that the campaign had run, that consumers had seen it, and that it had produced measurable results. In an awards culture that runs on case study videos and claimed impact metrics, this is where the deception lives.
The ghost campaign problem is older than AI
The coverage of the Cannes Lions scandal has focused, understandably, on AI as the enabling technology. Generative tools can now produce plausible-looking footage, fabricate data visualizations, and simulate broadcast appearances at a cost and speed that would have been prohibitive five years ago. The AI framing is accurate. It is also insufficient.
Ghost campaigns — entries that represent work that was never broadcast, never ran as described, or ran in a form significantly different from the submission — are not a product of the generative AI era. Industry insiders have discussed the practice informally for years. The incentives have always been present: Cannes Lions awards carry commercial value in the form of agency new business, talent recruitment, and client confidence. Before AI, fabrication required more effort: staged photographs, constructed media plans, invented metrics that couldn’t be easily verified. The work was harder to fake convincingly. AI has lowered the production cost of plausible-looking evidence to near zero. It didn’t create the incentive to deceive. It removed the most significant practical obstacle to acting on it.
The LePub São Paulo situation, which emerged alongside the DM9 scandal, illustrated this clearly. A Bronze Lion campaign for New Balance and São Paulo FC reportedly featured unverifiable sales data, and New Balance stated publicly that it had not approved the submission and had not signed off on the claimed results. This required no AI at all. It required an agency entering work the client hadn’t approved and attaching data the client couldn’t verify — practices that are structurally enabled by a submission process that, until this year, relied on professional trust rather than verification.
What the new rules require
Cannes Lions’ response to the scandal was substantial. The new integrity standards, taking effect for 2026 submissions, require that every entry be personally signed off by the business leader of the submitting agency and a senior marketer from the brand. This is the structural change with the most practical consequence: it places accountability at the executive level, making an agency CEO whose name appears on a fraudulent entry personally responsible for the submission in a way that a junior strategist or production team member previously was not.
The festival has also introduced a dual-layer verification system combining AI-detection tools and human review, an Integrity Council for escalated cases with independent oversight, and the possibility of three-year bans for agencies found to have submitted deliberately misleading work, with jury eligibility also subject to revocation. An AI Integrity Handbook, drawing on frameworks from the OECD and UNESCO, is being developed to define the boundaries of legitimate AI use in awards submissions.
These are meaningful reforms. They are also, implicitly, an admission that for years the festival had been operating without a credible verification mechanism — relying on professional honor and reputational incentive to deter fraud, both of which had clearly become insufficient as the commercial value of a Cannes Grand Prix increased and the tools available for constructing false evidence became more accessible.
The accountability question the rules don’t answer
The CEO and CMO sign-off requirement changes the incentive structure by relocating risk to people with more to lose. This is, in most institutional contexts, the operational definition of accountability: the decision-maker and the consequence-bearer are the same person. In the previous system, the person who decided to fabricate case study materials was unlikely to be the person who lost the most when the fabrication was discovered.
What the reforms don’t address — and what the industry should be asking — is the awards culture that made the incentive to fabricate so strong in the first place. Cannes Lions is not merely a celebration of good work. It is, for a significant portion of the agencies that enter it, a commercial necessity: a source of new business proof points, a factor in client conversations about agency quality, a driver of talent acquisition and retention. When the award carries that much commercial weight, the pressure to win it distorts the behavior of people who might otherwise not consider fabricating anything.
The festival cannot fully resolve this on its own. The industry would have to collectively decide that Cannes Lions results carry less weight as a commercial signal than they currently do — a decision that would require the clients who use them as a hiring criterion to change their behavior first. That is unlikely to happen quickly.
What can happen more quickly is what the new rules attempt: making the cost of fraud high enough that the calculation changes for the people who make it. Whether three-year bans and executive sign-offs are sufficient disincentives against Grand Prix-level commercial reward is, ultimately, an empirical question. The industry will find out over the next few festival cycles whether it has actually changed the system or whether it has changed the paperwork.
