Cannes Lions AI awards force agencies to prove creative originality

At the 2026 Cannes Lions festival in Cannes, France, running June 22-26, entrants must state whether AI was used in the work or the entry materials and explain the purpose of that use. For agencies trying to grow into enterprise accounts, the same controls that satisfy awards scrutiny also help protect creative credibility, client trust, and margin.

Why Cannes is setting the new baseline

As the largest gathering in the creative marketing community, Cannes Lions gives its rules outsized influence. In 2026, the festival added a Creative Brand Lion, which broadens the signal beyond individual campaigns and toward the systems that make strong work possible inside an organisation.

Entries may be ineligible if they cannot meet the stated requirements. The judging process also includes multiple reviews before an entry reaches the jury, and every entry is submitted to a specially built internal AI model for review. That mirrors the structured review gates agencies already use: human judgment is still central, and machine-assisted checking is part of the compliance stack.

The workflow agencies should build before AI scales across accounts

The first rule is simple: do not let AI touch client work without an approval path. A usable workflow starts with a named owner for the concept, a second owner for execution, and a final approver who can reject work that feels derivative, unverifiable, or strategically off brief.

The second rule is to separate ideation from final authorship. If a team uses AI for exploration, it should record what the tool contributed, what the human team selected, and what changed before delivery. That distinction matters because Cannes is asking not only whether AI was used, but why it was used, which means agencies need a clean record of human judgment at each stage.

A practical approval chain for client work looks like this:

That approval chain lets agencies scale AI across client accounts without creating a blind spot between the team that made the work and the team that owns the risk.

Disclosure standards should be written before the first prompt

Cannes Lions moved from optional disclosure in 2024 to mandatory disclosure after 2025. Agencies should not wait for clients to ask whether AI was involved, because by then the process has already become defensive. A better standard is to disclose AI use at the point of internal review and preserve that disclosure in the project record, pitch deck, and final submission copy where relevant.

Disclosure also needs categories, not just yes-or-no language. A useful agency standard should distinguish between AI used for:

• brainstorming or concept exploration
• drafting or rewriting copy
• image or video generation
• retouching, cleanup, or synthesis
• compliance review or quality control

That level of detail prevents the common trap where a team says it used AI, but no one can tell whether the tool shaped the idea or merely accelerated production.

Authorship has become a client-facing issue, not just a creative one

The Cannes rules put originality at the center of validation, and that changes how agencies should talk about authorship. If a campaign depends on a human insight, a distinctive point of view, or a specific strategic choice, that human contribution needs to be documented. If the machine generated the first draft but the team transformed it into something with a clear voice and purpose, that transformation needs to be visible internally.

Clients paying for strategic counsel do not want a black box that pumps out assets at scale. They want a team that can explain why a piece of work exists, who shaped it, and why the final version still carries a point of view that belongs to the brand.

Risk controls need to cover more than copyright

The 2025 withdrawal of DM9’s “Efficient way to pay” campaign over AI use and other manipulation concerns is the cautionary example here. It shows that the risk is not just whether AI was used, but whether the final work can withstand scrutiny around truthfulness, representation, and editorial integrity. Agencies need controls that catch manipulated imagery, misleading edits, unsupported claims, and reused material before the work leaves the building.

Cannes Lions’ tightened proof-of-impact requirements and mandatory executive endorsements point in the same direction. In 2026, Phil Thomas said 40% of entries disclosed AI use, up from 20% in 2025 and 11% in 2024, while total entries fell 25% to 20,050 from 26,900 the year before. Agencies that can document their process will be better positioned when a client asks how an idea was made, how a claim was checked, or who signed off on the final version.