Unilever AI Strategy: A Business Strategy Powered by AI

Unilever builds AI into its business strategy instead of running it as a separate program. AI’s first job is demand generation: faster marketing content, and brands that show up well when shoppers ask AI assistants. Unilever also has the most detailed AI governance process in this index so far.

There is no separate AI strategy. Allen Bates, Unilever’s Chief Digital and Information Officer for North America, said in October 2026: “We don’t have an AI strategy; we have a business strategy that’s powered by AI” (CIO Dive).

The ambition covers the whole company. CEO Fernando Fernandez said Unilever is “making our organization fit for the AI age, transforming every link in the value chain” (CIO Dive, February 2026). The 2025 annual report makes “Fit for AI Age” one of three shifts in Unilever’s strategy, with AI and technology “powering creativity, growth and margin expansion throughout our business” (Form 20-F, March 2026).

Unilever names the industry shift. Fernandez expects that “the consumer journey of discovery and shopping will be more conversational.” Chief Supply Chain and Operations Officer Willem Uijen added: “As brands are increasingly discovered and chosen in environments shaped by AI, we must lead this shift.”

The focus starts with demand. Unilever’s AI goals are to generate demand, innovate faster and improve operations:

• marketing: AI content studios, product “digital twins” for imagery, and Sketch Pro, an in-house generative AI design unit
• AI discovery: managing how its brands rank in large language models
• R&D: modeling and simulation to shorten innovation cycles
• manufacturing: digital twins of factories

The marketing bet is social-first. Fernandez has said he aims to put about half of Unilever’s marketing budget into social media (Marketing Dive), and AI-made content is what makes that volume possible. Unilever’s AI marketing work is one of the examples in What Real AI Leadership Looks Like.

Unilever has not said what it will not prioritize.

Google Cloud is the new foundation. In February 2026, Unilever signed a five-year partnership with Google Cloud to move enterprise applications and data platforms and build an “AI-first” foundation on Vertex AI and Gemini, including agentic workflows across business processes.

Other partners serve specific jobs:

• Microsoft: agents and other frontier capabilities
• Nvidia: Omniverse, which powers the product digital twins
• Accenture: more than 40 new factory digital twins over 18 months (The Manufacturer, June 2026)

Data comes first. Bates puts it this way: “AI loves context, context loves data, that’s where we start.” Unilever builds reusable data products so teams do not start from scratch.

Enablement goes wide and deep. In 2024, Unilever adopted a “go wide and go deep” AI approach. It has trained more than 23,000 employees on generative AI tools and runs more than 500 AI projects, alongside specialist units such as Sketch Pro.

Change is treated as everyone’s job. Bates says solutions are built by multidisciplinary teams, and warns that “technology on a bad process is still a bad process.” That is the core of AI change management: adoption is a people problem before it is a technology one.

The workforce stance is not stated. Unilever completed an €800 million productivity program in 2026, which included about 7,500 role cuts announced in 2024, but it was not presented as an AI program. Unilever has not said how AI will affect headcount.

Ownership is spread. Fernandez frames AI as company-wide, Uijen fronted the Google Cloud deal, and data and digital leaders run delivery. No single AI owner or AI budget has been disclosed.

Governance is the strongest in the index so far. Unilever has run an AI assurance process since 2021, with external partner Holistic AI (MIT Sloan Management Review):

• Proposed AI uses are reviewed before they are built.
• Each is scored for bias, privacy, robustness, explainability and efficacy, and rated red, yellow or green.
• A senior board with legal, HR, and data and technology members makes final calls.

One policy line sums it up: “We will never blame the system; there must be a Unilever owner accountable.” By July 2024, 150 projects had been assured, and Unilever had published how it is preparing for the EU AI Act.

AI fluency, coded against the Lead with AI Fluency Matrix (September 2026 edition):

• Fluency floor: Not disclosed. More than 23,000 employees have been trained, but no mandate or standard was found.
• Highest level evidenced: Level 3 (Integrating), company-claimed. Marketing teams produce content with AI studios and digital twins, with reported speed and cost results. Agentic workflows are planned with Google, but no results have been published, so Level 4 is not evidenced.
• Portfolio design: Wide and deep. Broad training sits alongside specialist AI units and hundreds of projects. No role-specific targets have been disclosed.

Unilever reports results by use case. All figures below are company-claimed.

Marketing content:

• Digital twins produce product imagery about twice as fast at roughly half the cost (March 2025).
• Its Beauty AI Studio produced assets up to 30% faster (October 2026).

Manufacturing: At its Raeford, North Carolina plant, a digital twin raised capacity by 10%, cut waste by 20% and predicts 95% of process flow restrictions in deodorant stick production (June 2026).

AI discovery: By July 2026, Unilever was managing its brands’ rankings in large language models in more than 20 markets for more than 20 brands, Fernandez said.

Two things are missing: an AI value figure and AI targets.

The roadmap runs through the Google Cloud partnership: more than 40 new factory digital twins over 18 months, and agentic workflows across the business over five years.

Change log

• 2021: The AI assurance process is formalized with Holistic AI.
• 2024: Unilever adopts a “go wide and go deep” AI approach.
• July 2024: 150 AI projects have been assured, and Unilever publishes its EU AI Act preparation.
• March 2025: Digital twins are reported to halve imagery cost and double speed.
• 2025: Sketch Pro launches, more than 23,000 employees are trained and more than 500 AI projects are running.
• February 2026: A five-year Google Cloud partnership is announced.
• June 2026: Raeford digital twin results are published, and more than 40 new digital twins are planned.
• July 2026: Brand rankings in large language models are managed in more than 20 markets.
• October 2026: “We don’t have an AI strategy; we have a business strategy that’s powered by AI.”

• Disclosure: 4 of 4. Unilever answers all four questions, though its measurement is by use case, with no value figure.
• It has the most detailed AI governance in the index. Every AI use is reviewed and rated, and a named person stays accountable.
• AI is part of the business strategy, not a side program, with demand generation first.
• It is preparing for AI-led discovery by managing how its brands rank in AI assistants.
• The say-do gap is people and value. Training is broad but has no standard, the workforce stance is unstated, and there is no AI value figure.
• Watch the Google Cloud build. The test is whether agentic workflows produce a value figure Unilever is willing to publish.

What the numbers show: On a comparable basis, headcount fell 7% while revenue slipped 2%, lifting revenue per employee 6%. That period overlaps Unilever’s productivity program, which cut about 7,500 roles and was not presented as an AI program.

Revenue, headcount and revenue per employee, FY2022 to FY2025

• FY2022: €60.07bn revenue, 138,000 employees, €435k per employee
• FY2023: €51.68bn revenue, 108,000 employees, €479k per employee
• FY2024: €52.48bn revenue, 105,000 employees, €500k per employee
• FY2025: €50.50bn revenue, 100,000 employees, €505k per employee

Change from FY2023, the first comparable year, to FY2025: revenue down 2%, headcount down 7%, revenue per employee up 6%.

Notes: Revenue is turnover, as reported. Headcount is the average number of employees during the year. 2022 is as reported and includes the Ice Cream business, demerged in December 2025. 2023 to 2025 are restated to exclude Ice Cream, so the comparable trend starts in FY2023. Trend figures stay in € so exchange-rate moves do not distort them. At the 2025 average exchange rate, latest revenue per employee is about $571k.

These figures give context on scale and productivity. They do not show that AI caused any change. Sources: Unilever 20-F filings.

What is Unilever’s AI strategy?

Unilever treats AI as part of its business strategy rather than as a separate program. It uses AI to generate demand through faster marketing content, prepare its brands for AI-led shopping, speed up R&D and improve manufacturing. A five-year Google Cloud partnership is its technology foundation.

‍

How does Unilever use AI in marketing?

Unilever uses AI content studios and product digital twins to produce marketing content faster and at lower cost. It says digital twins produce imagery about twice as fast at about half the cost. It also manages how its brands rank in AI assistants in more than 20 markets.

‍

How does Unilever govern AI?

Since 2021, Unilever has reviewed proposed AI uses through an assurance process with its partner Holistic AI. Each use is scored for risks such as bias and privacy and rated red, yellow or green, and a senior board makes final decisions.

‍

Is Unilever cutting jobs because of AI?

Unilever has not linked job cuts to AI. It completed a productivity program in 2026 that included about 7,500 role cuts announced in 2024, but that program was not presented as an AI initiative.

‍

How is this profile put together?

Lead with AI profiles use public sources only, such as annual reports, earnings materials, filings and on-record interviews, dated at the time of writing. AI fluency is coded against the Lead with AI Fluency Matrix. Company figures are marked as company-claimed unless they have been independently verified.