Three levers, and the one the CFO will test
The case rests on three levers. Cost: production and adaptation spend, agency fees for versioning and localisation, and internal hours. Speed: weeks from brief to market, which matters most where launches, seasons and competitors set the clock. Growth: more ideas tested and more markets served from the same budget. CFOs test the first lever hardest, so cost one recent campaign end to end, including review time, and compare a pilot on similar work costed the same way. Count new costs openly (platform fees, usage, the review time that grows with volume) and present savings as reinvestment, not as a headcount promise made before the evidence.
Where the organisation actually is
It helps to be honest about the starting point. A simple maturity model has four stages. First, scattered experiments with free tools, nobody accountable and unknown risk. Second, a sanctioned platform for some teams, with basic governance. Third, AI built into the content supply chain, with briefs, brand rules, approvals, DAM and channels connected and production measured end to end. Fourth, content planned around evidence of what works for each audience and market. The investment and the risks differ at each stage. Jumping from the first stage straight to scale is how brands end up with thousands of assets nobody can trace.
Risks the board will ask about
Expect three questions. Brand: can an off-brand or offensive asset reach the market, and who would catch it? Legal: are the models cleared for commercial use, are disclosure duties such as the EU AI Act's rules on deepfakes (from 2 August 2026) being met, and is there an audit trail? Data: could unreleased products or customer information leak through a vendor? Answer each with a control rather than a promise. In Synthetic White those controls are part of the platform: brand rules and approved claims enforced at generation, approvals with an audit trail, model provenance logged for each asset, zero-retention agreements and no training on customer uploads.
Rethinking the agency roster
The largest organisational change often sits outside the building. Decide which work stays with agencies (strategy, big ideas, flagship productions) and which moves to generation, whether in-house or inside a shared platform: adaptation, versioning, localisation and routine production. Re-scope contracts around outputs and ideas instead of billed adaptation hours, and time the change with renewals instead of springing it mid-year. Ask each agency where it already uses AI on your account and how that efficiency shows up in its fees. Good agencies tend to welcome the shift, because it moves their paid time towards the thinking clients value most.
Updated 25 September 2026