AI content operations aren't a hypothesis. Here's what documented, cited results from brands actually running them look like, the wins and the misses.
Industry-wide data on how fast AI content production is being adopted, and what it's actually changing.
Three wins and one honest miss. The difference between them was never access to AI. It was craft judgment, process and who was in the room, which is exactly the gap our services and talent are built to close.
Heinz prompted DALL-E 2 with phrases like "ketchup bottle stained glass" and "ketchup in psychedelic art." The AI kept drawing Heinz's distinctive bottle shape without being asked to, and Heinz turned that into the campaign: proof that their brand is so iconic even an AI trained on the whole internet defaults to it.
Director Gibson Hazard's team used Midjourney extensively during preproduction to explore visual direction for Nike's Travis Scott spot, generating a wide range of concept imagery before committing to a final look. The finished ad was then built with a hybrid workflow: AI-assisted exploration, traditional VFX execution.
Since 2023, WPP has been building a generative-AI content engine with NVIDIA Omniverse, connecting 3D design, Adobe and Getty Images tools so creative teams can produce client-ready commercial content at a fraction of the previous time. By 2025–2026 the partnership had expanded into agentic AI and robotics-assisted production.
For two years running, Coca-Cola released AI-generated "Holidays Are Coming" ads at real technical scale, one year's version reportedly involving around 100 staff producing roughly 70,000 AI-generated clips over 30 days. Both years drew significant public backlash over the "AI sheen" and a perceived lack of real craft.
Hiring for something like the wins above, or the person who could produce them? Either way, start here.