

Brand governance when the machine can make everything: what IKEA's CMO said, and what nobody has documented
ContentGrip reports that IKEA's Global CMO sees AI as a companion, not a replacement for creative. That is worth hearing. It is also, as far as the record shows, a stance and not a published rulebook, which makes it a useful test of what brand governance really means.
By Katie Delaney / 2026-10-15 / 8 min read

What IKEA's CMO said about AI, and the limits of the record on brand governance
The fox weighs the source before the slogan. On 15 October 2026 ContentGrip wrote that IKEA is treating artificial intelligence as a way to carry its brand further, not as a system that gets to decide what the brand means. It reports that Global CMO Vincenzo Riili says the retailer is already using AI to adapt work faster across markets and that he does not view marketing creative as ready for full automation, framing AI as a companion and warning that overreliance could weaken IKEA's distinctive way of appearing in the market.
Now the careful part. The ContentGrip article states that it is produced by ContentGrow, it links no IKEA press release or governance document, and its headline word boundaries is the publisher's own framing. So the claim that IKEA has written limits on AI marketing scale is not supported by a primary document in the sources read for this piece. What is on the record is a stance, which is useful, and it is a CMO's, which is still opinion about future conduct.
The closest IKEA-side primary is an Ingka Group page carrying a transcript of Riili's appearance on a podcast. It records him saying We're not ready yet to fully compete with human creativity, but we're not far, and, on the commercial stake, if someone asks an AI for the best value sofa and IKEA doesn't appear in the response, there's a problem. Neither line describes an approvals process, and the page carries no figures on AI content volume.
An earlier IKEA voice adds texture. In a June interview with Euronews, Dorothée Gardon, IKEA's Global Marketing Strategy and Integrated Media Director, said that AI is wonderful at creating valuation, but it cannot replace the intelligence that comes from the human brain and that the company must stay in control of brand safety because hallucinations do happen. Two executives, one stance: AI helps, and people keep the pen.
Brand governance is a distinctiveness problem, not a content problem
If the stance is clear and the rulebook is not, what should brand governance contain? Start from what brands are protecting. The Ehrenberg-Bass Institute's Julian Major put the rule in one line: consistently use your distinctive assets: do not change from campaign to campaign, and be wary of refreshing elements. Generative tools make change cheap, and that is the risk. When a model can produce a thousand variants in an afternoon, restraint becomes the scarce skill.
WARC's September 2026 summary of the same idea reads: Leaders run brand as an operating system, one source of truth, with AI held as accountable as everyone else. Gartner's 2025 work on training AI for on-brand content is blunter about the practical gap: gaps remain in their ability to generate on-brand, commercially publishable media at scale, and it says detailed brand guidelines design principles must be consistently applied.
That last phrase is the working end of brand voice guidelines. A document written for human freelancers is rarely precise enough for a model. A set of AI brand guidelines needs examples a machine can imitate, counter-examples it must avoid, and a named person who can say no. Good brand guidelines examples show the rule, the exception and the reason, because a model, like a new hire, needs the reason.
A practical note on language. An AI brand voice is not a tone adjective list. It is a bundle of choices: sentence length, what you never say, which claims need proof, and what a regional team may adapt. IKEA's reported use, adapting work faster across markets, is exactly the case where a guardrail matters, because local change multiplies every ambiguity in the master.

What customers think, and what regulators already expect of AI brand guidelines
Governance is not only an internal matter. Gartner's March 2026 survey of 1,539 US consumers, conducted in October 2025, found that half of U.S. consumers (50%) say they would prefer to give their business to brands that do not use GenAI. The same release says 68% frequently wonder whether the content and information they see is real and sixty-one percent (61%) of consumers say they frequently question whether the information they use is reliable. Its advice to marketers includes to label AI-driven experiences, so people understand when and how AI is being used.
| Item | Value |
|---|---|
| Wonder if content is real | 68 |
| Question information | 61 |
| Prefer no GenAI brands | 50 |
The rulebook outside the building is moving too. The UK's Advertising Standards Authority said in June that the Code is media-neutral, meaning the rules apply regardless of how the content was created and that if an AI tool produces something problematic, the advertiser remains responsible for any output used in their ads. In February the Advertising Association published a voluntary guide built on eight principles for responsible use.
The EU AI Act adds a transparency layer. Article 50 requires that outputs of AI systems generating synthetic content are marked in a machine-readable format and detectable as artificially generated or manipulated, with separate disclosure duties for some published text. Whether a given marketing asset falls inside those duties is a legal question, but a brand that already logs which assets were AI-made is halfway to answering it.
Five rules of brand governance a team can adopt this month
These rules are folkfox's own synthesis of the evidence above, not IKEA's. They fit alongside our brand strategy work and echo recent folkfox pieces on brand architecture, brand salience and shopping agents and how holding names whisper.
Give brand governance a single accountable person, with authority to refuse an asset, and publish their name internally.
Turn brand voice guidelines into examples, counter-examples and reasons, so a machine and a new hire read the same page.
Let AI adapt, translate and resize, and keep the brand idea and the first creative decision with people.
Record which assets used generative tools, so disclosure, claims review and rights checks have a trail.
Check each batch against the brand's distinctive assets before release, and retire variants that dilute them.
Perhaps the most vulpine thing IKEA's CMO said is the least dramatic: that the aim is to carry the brand further, not to let a model decide what it means. A fox does not guard every field, only the den. Keep the few things that are unmistakably yours under a steady hand, and let the machine run along the hedgerow.
Frequently asked questions
What is brand governance?
Brand governance is the set of roles, rules and approvals that keep a brand consistent as more people and tools produce its content. It names an owner, a source of truth and a review step. With generative AI it also records which assets were machine-made.
What should AI brand guidelines include?
AI brand guidelines should include examples a model can imitate, counter-examples to avoid, banned claims, tone choices with reasons and a named approver. A list of adjectives is not enough. Gartner's 2025 work says detailed brand guidelines and design principles must be consistently applied.
What are good brand guidelines examples?
Good brand guidelines examples show a rule, an exception and the reason in the same place, such as how a logo may be used and a case where it may not. For AI use, add prompts that worked, outputs that were rejected and why.
How do brand voice guidelines change with AI?
They need to be more precise. A tone description suits a human freelancer, but a model needs sentence-level examples, forbidden phrases, claim rules and regional adaptation limits. Treat the guidelines as a working specification that a person owns and updates.
Did IKEA publish rules limiting AI in marketing?
No published IKEA rulebook was found in the sources read for this article. ContentGrip reports the CMO's stance that AI is a companion and creative is not ready for full automation. The word boundaries in its headline is the publisher's framing, not an IKEA document.
Who is responsible for AI-made adverts in the UK?
The advertiser. The ASA said in June 2026 that the Code is media-neutral and that if an AI tool produces something problematic, the advertiser remains responsible for any output used in its ads. That applies whether a person or a model drafted the copy.
Brand salience and the shopping agent that, reportedly, picks famous first
Why distinctiveness matters when an agent chooses.
Read the pieceBrand strategyIntuit's One Intuit reboot, and the brand architecture examples worth copying
How a large brand organises many names.
Read the pieceBrand strategyCorporate rebrand: why holding names whisper while consumer names sell
Which brand decisions deserve the loudest guard.
Read the pieceBrand strategyA brand portfolio strategy is a set of jobs, and BT just relit the oldest lantern
Roles for each brand in a portfolio.
Read the pieceDoes your brand have a rulebook a model can read?
folkfox writes brand governance for the AI era: an owner, model-ready guidelines and a review step that keeps scale from outrunning what makes you recognisable.
Want folkfox in your Google results and AI answers? Set folkfox as a preferred source.

