When a brand the size of adidas makes a technology bet, the whole industry watches. The question worth asking is not whether adidas is investing in AI — it clearly is — but how it is structuring those investments. The answer, visible across its publicly disclosed moves, points to a deliberate open-innovation posture: partner broadly, build selectively, and never let a single vendor own the keys to your creative or commercial infrastructure.
Key takeaways
- adidas treats AI partnerships as a portfolio, not a dependency — spreading capability across external vendors and internal teams simultaneously.
- Open innovation in sportswear is partly a hedge: brands that rely on a single AI platform risk losing negotiating power as those platforms mature and reprice.
- The fashion industry at large is navigating a period of low growth, making cost-efficient digital product creation a strategic priority rather than a nice-to-have.
- For fashion-tech founders pitching enterprise deals, adidas's model signals that flexibility, interoperability, and data sovereignty matter more to large brands than any single feature.
- Vogue Business and other trade titles have tracked how AI is reshaping the brand-technology relationship — the conversation has moved from 'if' to 'how fast' and 'on whose terms.'
What does 'open innovation' actually mean for a sportswear giant?
Open innovation, in its simplest form, means a company deliberately draws on external ideas, tools, and partners rather than assuming everything valuable must be built inside. For a brand like adidas, that means running AI pilots with outside vendors, licensing specialist tools, and sometimes acquiring capabilities — while simultaneously investing in proprietary data assets and internal engineering.
This is not the same as outsourcing strategy. The distinction matters: outsourcing hands control to a third party; open innovation keeps the brand at the centre, curating a constellation of partners it can reconfigure as the technology shifts. The goal is optionality.
In sportswear, where product cycles are short and trend windows are measured in weeks, that optionality has real commercial value. A brand locked into a single AI platform for design generation, demand forecasting, or digital sampling cannot easily pivot when a better tool emerges — or when the incumbent raises prices after achieving lock-in.
How does adidas structure its AI partnerships?
adidas has publicly signalled investment across several AI-adjacent domains: generative design, digital product creation, consumer personalisation, and supply-chain optimisation. Rather than building a single monolithic AI platform, the brand appears to treat each domain as a separate capability question — and sources answers from different partners.
This approach mirrors what Vogue Business has documented across the luxury and sportswear sectors: leading brands are assembling best-of-breed stacks rather than committing to one vendor's end-to-end vision. The logic is sound. AI capabilities are evolving fast enough that a tool that leads today may be commoditised within two product cycles. Keeping partnerships modular means the brand can swap components without rebuilding from scratch.
For fashion-tech founders, this is the single most important structural insight: adidas is not looking for a platform that does everything. It is looking for partners that do one thing exceptionally well, integrate cleanly, and do not demand exclusive data relationships.
Why avoid vendor lock-in now?
The timing is not accidental. The fashion industry is operating in a period of constrained growth — McKinsey's State of Fashion research notes that low growth is expected to persist across fashion and luxury retail globally, with macroeconomic headwinds weighing on both Europe and key export markets. In that environment, brands are scrutinising every technology contract more carefully than they did during the expansion years.
Vendor lock-in carries a specific risk in this climate: if an AI platform becomes essential to your product-creation workflow and then reprices, you either absorb the cost or face an expensive migration. adidas's open-innovation posture is, among other things, a procurement strategy — one that keeps multiple vendors competing for its business rather than any single one extracting monopoly rents.
There is also a data-sovereignty dimension. adidas's design archive, consumer behaviour data, and supply-chain intelligence are among its most valuable assets. Feeding those assets into a third-party AI platform that retains training rights is a very different proposition from using a tool that processes data in a controlled, brand-isolated environment and returns outputs without retaining the underlying IP.
What does adidas actually build in-house?
This is where the public record gets thinner, and honest analysis requires acknowledging that. adidas has spoken publicly about digital product creation — the ambition to design, sample, and approve products digitally before committing to physical production — and about using data to sharpen demand forecasting. But the precise boundary between what is built internally and what is licensed or partnered is not fully disclosed.
What is visible: adidas has invested in internal data infrastructure, employs significant engineering talent, and has run pilots in 3D design and digital sampling. The brand has also been active at industry forums — events like those run by PI Apparel, where brands and fashion-tech vendors meet to evaluate tools — which signals active market scanning rather than passive vendor selection.
The in-house build question is ultimately about where the brand believes it has durable competitive advantage. Proprietary consumer data and brand aesthetic judgment are hard to replicate; the AI tooling that processes them is increasingly commoditised. adidas appears to be investing in the former while treating the latter as a procurement decision.
What does this mean for fashion-tech founders pitching enterprise deals?
If you are building a tool and hoping to land adidas or a brand of comparable scale, the open-innovation model gives you a clearer brief than most enterprise sales processes do.
First, interoperability is non-negotiable. Your tool needs to connect to existing workflows — PLM systems, 3D design environments, supply-chain platforms — without requiring the brand to rebuild around you. If your pitch begins with 'replace your current stack,' you have already lost the room.
Second, data sovereignty is a dealbreaker at the top end. Enterprise brands will ask, directly or through procurement, what happens to their data. If your model trains on customer inputs or retains design files, that is a structural objection you need to answer before it is raised.
Third, prove the module, not the vision. adidas's modular approach means it is evaluating tools against specific capability gaps, not broad platform promises. A founder who can demonstrate a measurable improvement in one domain — digital sampling speed, demand-signal accuracy, colourway generation time — is more credible than one who promises to transform the entire product-creation process.
Fourth, be honest about where you sit in the stack. Brands at this scale have seen enough AI pitches to recognise when a vendor is overselling scope. Knowing your lane — and being able to articulate how you complement the tools the brand already uses — is a competitive advantage in itself.
How is the broader industry reading adidas's posture?
The open-innovation model is not unique to adidas, but the brand's scale makes its choices legible as a signal for the rest of the market. When a brand with adidas's resources chooses breadth over depth in its AI partnerships, it tells smaller brands that the technology is not yet mature enough to justify a single-vendor bet — and it tells vendors that the enterprise market will reward specialisation and interoperability over platform ambition.
Vogue Business has been tracking how AI is reshaping the brand-technology relationship across fashion and luxury, noting that the conversation among executives has shifted from scepticism to implementation — the live questions now are about governance, data rights, and which capabilities are genuinely differentiating versus which are becoming table stakes.
For consumers, the downstream effect of this innovation architecture is products that are more precisely matched to demand, with shorter gaps between trend signal and shelf availability. Digital product creation, when it works, reduces the sample waste and overproduction that have long been structural costs of the fashion industry. That is not a small thing in a period of sustained low growth.
What should you watch next?
The most useful signals to track in adidas's AI strategy over the coming product cycles are:
- Digital product creation adoption rate: how much of the line is designed and approved digitally before physical sampling begins. This is the metric that most directly captures the ROI of AI investment in product development.
- Partnership announcements vs. acquisition moves: open innovation tends to shift toward acquisition when a capability becomes truly core. If adidas moves from partnership to ownership in a specific domain, that domain has crossed from 'commodity tool' to 'strategic asset' in the brand's own assessment.
- Data governance disclosures: as AI regulation tightens in the EU and elsewhere, brands will be required to be more explicit about how they use AI in product decisions. adidas's disclosures in this area will reveal the actual shape of its AI stack more clearly than any press release.
- Performance at industry forums: events like PI Apparel, where brands and vendors meet in structured formats, often surface the capability gaps that brands are actively trying to fill — and the vendors they are evaluating to fill them.
The open-innovation model is not a permanent state. It is a posture appropriate to a moment when AI capabilities are advancing faster than any single vendor can track. When the technology stabilises — when the tools that matter become clear — expect the portfolio to consolidate. The brands that have kept their options open will be better placed to make that consolidation on their own terms.
FAQ
What is open innovation in the context of a fashion brand? Open innovation means a brand deliberately sources ideas, tools, and capabilities from external partners rather than building everything internally. For fashion brands, this typically means partnering with AI vendors for specific functions while retaining control of core data and brand judgment.
Why is adidas cautious about vendor lock-in? Lock-in gives a single vendor pricing power once a tool becomes embedded in core workflows. In a low-growth market, that is a structural cost risk. adidas's modular partnership approach keeps multiple vendors competing for its business.
What do fashion-tech founders need to know when pitching enterprise brands? Interoperability, data sovereignty, and a focused value proposition in one domain matter more than platform breadth. Enterprise brands are evaluating specific capability gaps, not end-to-end replacements.
How does digital product creation reduce costs for a sportswear brand? By designing and approving products digitally before committing to physical samples, brands cut material waste, shorten approval cycles, and reduce the risk of overproduction — all of which matter more when growth is constrained.
Where can I follow coverage of AI in fashion brand strategy? Vogue Business publishes dedicated editorial series on AI and technology across fashion and luxury, tracking how brands are implementing and governing AI tools at scale.