# Coca-Cola replaces José Mourinho with AI clone in campaign, cutting celebrity partnership costs *The beverage giant's AI-generated talent play signals a new route around exclusive athlete contracts and scheduling complexity.* By **Jenny Huang Goodman MPA MSc MHSA, Principal** — The Stash Edge, Hako Shikin LLC. Published 2026-06-06. Canonical: https://www.pops4.com/stash/articles/coca-cola-2026-06-06t03-6 Subject: Coca-Cola Tags: ai talent, celebrity licensing, content production, cost reduction, video marketing, influencer replacement --- Coca-Cola's upcoming campaign for Real Madrid will feature an AI-generated version of manager José Mourinho, not the manager himself, according to Digiday. The brand commissioned a digital clone that replicates Mourinho's voice, mannerisms, and likeness, then deployed it across video creative without negotiating appearance fees, travel logistics, or shooting schedules with the actual person. The mechanics are straightforward. Coca-Cola licensed Mourinho's digital rights, trained a generative model on existing footage and voice recordings, then scripted and rendered the campaign content entirely in-house. The AI clone delivers scripted lines in multiple languages, appears in contexts Mourinho's calendar would not accommodate, and requires no physical presence. Digiday reports the brand views this as a test case for scaling celebrity partnerships without the cost structure or scheduling friction of traditional talent contracts. This works because the value in celebrity endorsement has migrated from physical presence to recognizable identity. Consumers respond to the visual and tonal signature of a known figure, not the metaphysics of whether that figure stood on set. An AI Mourinho reading copy in Portuguese, Spanish, and English delivers the same brand association as the man himself, but at a fraction of the cost and with zero coordination overhead. The brand retains control over message, timing, and distribution, and the talent receives a licensing fee instead of a per-appearance rate. For a small physical-product brand, the steal is this: license micro-celebrity or expert likenesses at fixed rates, then generate testimonial and unboxing content on demand. Identify a recognized figure in your category — a YouTube reviewer, a niche athlete, a regional designer — and negotiate a one-time digital rights deal for **$2,000 to $8,000**. Use accessible tools like Synthesia or HeyGen to clone voice and appearance, then script product commentary, how-to videos, or founder interviews that would otherwise require coordinating schedules and paying day rates. A candle brand could commission an AI version of a known fragrance blogger to narrate seasonal scent profiles. A knife maker could render an AI clone of a respected chef demonstrating edge retention. The content library scales without additional talent fees, and the brand controls release timing to match product drops or seasonal windows. The cost structure flips. Traditional talent partnerships for physical products run **$5,000 to $50,000** per campaign, plus travel, coordination, and approval rounds. AI licensing costs **$2,000 to $8,000** upfront, then near-zero marginal cost per asset. A $10,000 budget that once bought two influencer posts now funds ten AI-generated videos, each tailored to a different product SKU or regional audience. The brand also avoids the risk of talent controversy or contract renegotiation; the digital asset remains stable and reusable. The broader pattern is that celebrity value is unbundling. Brands no longer need the whole person — just the recognizable signals that drive association and trust. Coca-Cola's Mourinho play is a proof point: the campaign delivers brand lift without the logistics tax of coordinating a global sports figure. For smaller brands, the same mechanism applies at micro scale. License the identity, control the message, and ship content faster than competitors still negotiating talent calendars. ## The takeaway License micro-celebrity digital rights once, then generate product content on demand at near-zero marginal cost. --- ## Publisher **Hako Shikin LLC** — Virginia Beach, Virginia. Founded 1997. ASI 217876 · DUNS 18-204-6339. Principal and author: **Jenny Huang Goodman MPA MSc MHSA**. - Author: https://www.huanggoodman.com/about - LLM context: https://www.pops4.com/stash/llms.txt - MCP endpoint, for AI agents: https://mcp.pops4.com/mcp - Client dashboard: https://dashboard.pops4.com/ - Catalogue: 70,000+ products, 200+ brands