Unilever brand teams are using generative AI tools — Midjourney, DALL-E, and various video tools — for content exploration and concepting. Adoption is patchy: some global brand teams are genuinely AI-native in production, others haven't started. There's no shared tooling standard, no prompt library, and no quality baseline across markets. The capability exists. The system doesn't.

Unilever has strong first-party data infrastructure, particularly through loyalty and D2C channels. AI-driven personalisation is live in some markets — programmatic and CRM — but disconnected from brand teams. The data is there. The pipeline from data to personalised creative isn't.

Media agencies are running AI-powered optimisation on paid campaigns. This is effectively outsourced capability — Unilever's in-house marketing teams aren't directing it. Good outcomes, low understanding. If the media agency changes, the capability disappears.

Exploratory. No standardised creative AI toolkit exists across brand teams. Individual teams are experimenting. Nothing has been adopted at scale, codified, or shared. The gap between what's possible and what's actually in production is widest here.

Significant variance by market and seniority. Some senior brand managers are AI-native. Most are not. No structured AI literacy programme exists for the marketing function. What's been learned has been learned informally. This is the highest-leverage gap — raising the floor here multiplies the impact of every other investment.

Unilever's governance frameworks are among the strongest in the FMCG sector. Their Responsible Beauty commitments, anti-retouching standards, and sustainability AI work set precedent. The challenge is that these frameworks were built for a different AI context — they govern what AI can produce, not how the marketing team uses AI to produce it.