Verified
Unilever
Scaling- Consumer goods
- 96,092 employees
- AI returns in factories and marketing
- 40,000+ trained in AI (Jun 2026)
- Google Cloud deal (Feb 2026)
Unilever credits AI with returns in its factories and marketing, such as 1% to 2% ingredient savings at its Cu Chi plant and product images at half the cost, but reports no company-wide AI figure. Unilever has implemented more than 500 AI projects and in Feb 2026 announced a five-year deal to move its data and cloud platform to Google Cloud and build on Vertex AI and Gemini. Chief Digital & Technology Officer Sam Kini said in Jun 2026 that more than 40,000 employees had been trained to use AI.
unilever.comTheir approach to AI
- The platform
- Unilever builds AI into its own tools, such as Brand DNAi for brand-safe content and an R&D Assistant agent over 150,000 scientific documents, and in Feb 2026 chose Google Cloud's Vertex AI as the base of an enterprise data and AI backbone.
- The models
- Google Cloud said in Feb 2026 that it is deploying Gemini for Unilever; Unilever teams also build with Microsoft Copilot Studio, and Unilever has named no other model provider.
- The reach
- Unilever had trained more than 40,000 of its roughly 96,000 employees to use AI by Jun 2026, and more than 75% of office-based Personal Care staff used personal AI productivity tools regularly in Nov 2025.
- The catch
- Unilever reports AI returns for single factories and tools, not for the company, and has not given a count of employees using AI, named a company-wide AI assistant or published an AI budget.
Snapshot
- Employees
- 96,092
- Dec 31, 2025, down from 120,040 a year earlier, mainly from the Ice Cream demerger
- Technology budget
- Not disclosed
- The value of the Feb 2026 Google Cloud deal was not published
- AI/ML headcount
- Not disclosed
- 40,000+ staff trained in AI (Jun 2026); about 2,000 training to build agents (Sep 2026)
- Who owns AI
- Sam Kini, CDTO
- Global CIO until Jan 2026; reports to Willem Uijen, not on the Unilever Leadership Executive (2025 Form 20-F)
- Platform
- Google Cloud Vertex AI plus in-house tools
- Five-year partnership announced Feb 17, 2026
- Models
- Google Gemini
- Named by Google Cloud in Feb 2026; Unilever names no other provider
- Reach
- 40,000+ trained
- Jun 2026; no company-wide count of regular users published
- Use cases
- 500+ AI projects
- Implemented over a decade (Feb 2025); more than 330 active in Sep 2024
Key findings
Unilever reports AI returns factory by factory and tool by tool, not for the company: an AI-driven mixer saves 1% to 2% of premium ingredients at Cu Chi, Vietnam, an AI cleaning system saves about €100,000 a year on two Poznan, Poland lines, and product images made with digital twins and AI cost half as much; Unilever plans more than 40 AI digital twins across its factories over 18 months from Jun 2026 with Accenture.
Unilever had trained more than 40,000 employees to use AI by Jun 2026, up from about 16,000 in Sep 2024 and 23,000 at the end of 2024, against a workforce of 96,092 at Dec 31, 2025, but has not said how many employees use AI tools regularly.
Unilever announced a five-year partnership with Google Cloud on Feb 17, 2026 to move its integrated data and cloud platform and key enterprise applications to Google Cloud and build on Vertex AI and Gemini, nearly three years after it finished an 18-month migration to Microsoft Azure in Apr 2023.
Sam Kini, who joined Unilever in 2021 and became Global CIO, has been Chief Digital & Technology Officer since Jan 2026 and reports to supply chain and operations chief Willem Uijen, according to Unilever's 2025 Form 20-F; Unilever named Reema Jain CIO under Kini in Feb 2026, and neither Kini nor Jain sits on the Unilever Leadership Executive.
Unilever has reviewed AI projects through an AI assurance process since 2019, and about half of the roughly 150 projects it put through that review by Sep 2024 needed fixes for performance, bias or transparency; its 2025 annual report says three policies were updated for AI risks in intellectual property, data privacy and marketing.
Where Unilever sits
- Stage
- Scaling
Unilever credits AI with business returns in two functions: in manufacturing, an AI digital twin mixer saves 1% to 2% of premium ingredients at its Cu Chi, Vietnam factory (Jun 2026) and an AI cleaning system saves about €100,000 a year on two Poznan, Poland lines (May 2025); in marketing, product images made with AI-driven product twins cost 50% less (Mar 2025). Unilever gives no firm-level AI figure and does not credit its €800 million productivity programme to AI. Unilever has not reported how many employees use its AI tools: Chief Digital & Technology Officer Sam Kini, who owns the AI-first agenda, said in Jun 2026 that AI-powered digital twins are cutting energy use and optimising ingredients in Unilever's factories and that more than 40,000 employees had been trained to use AI, which counts training rather than use, and Unilever's only regular-use figure covers office-based Personal Care staff. Unilever is therefore at Scaling, at its low edge because its returns are site and tool level and its Google Cloud data and AI backbone is still being built; Unilever falls short of Embedded because it credits no firm-level cost, revenue or margin figure to AI and reports no weekly AI use by most of its 96,092 employees.
How the stages are defined - Ownership
- CDTO Sam Kini, reporting to supply chain and operations chief Willem Uijen
- Platform
- Hybrid: Google Cloud Vertex AI backbone plus in-house tools
- Models
- Google Gemini; other providers not named
- Hosting
- Microsoft Azure since a 2023 migration; data and key applications moving to Google Cloud from 2026; no own data centers named
- Reach
- 40,000+ employees trained in AI (Jun 2026)
How they deploy
AI-first as a transformation pillar
Unilever calls becoming AI-first one of the three foundational pillars of its business transformation, and Sam Kini said in Jun 2026 that Unilever applies AI to growth and the future of commerce, to an intelligent enterprise across the value chain, and to employees' everyday work. Unilever's 2025 annual report lists an organisation Fit for the AI Age among the three shifts behind its strategy, alongside Desire at Scale and a Play to Win culture.
Go wide and go deep
Unilever adopted a go wide and go deep approach in 2024: going wide means productivity gains across the whole workforce, and going deep means high-impact opportunities. Unilever told CIO Dive in Sep 2024 that it had more than 500 AI projects worldwide, more than 330 of them active, and Steve McCrystal, then Chief Enterprise Technology Officer, said in Feb 2025 that AI had been part of Unilever's systems for more than a decade.
Brands built for people and algorithms
Fernando Fernandez told the dbAccess Global Consumer Conference, in remarks reported on Jun 5, 2026, that AI favours scientific proof, patents, claims and ratings, and that Unilever tracks how more than 50 of its brands appear in large language models against competitors on Amazon in 20 geographies. Sam Kini said in Jun 2026 that Unilever's brands must be designed for both people and algorithms, and the Feb 2026 Google Cloud partnership includes building agentic commerce and marketing capabilities.
Digital science in R&D
Fernando Fernandez said in remarks reported on Jun 5, 2026 that 80% of the work in Unilever's laboratories is now digital, against 20% five years earlier, and that innovation timelines have fallen from two to three years to nine to 12 months. Unilever says it can run more than 10,000 virtual experiments by connecting more than 25 million consumer and laboratory data points, and its Beauty & Wellbeing scientists use AI tools over Unilever's own R&D databases.
Process redesign before tools
Vicky Cuthbert, Unilever's Chief Product Supply Chain Officer for Personal Care, said in Aug 2026 that the larger opportunity lies in redesigning processes so people and AI work together, and that Unilever is moving from AI in individual tools to connected systems. Graham Sommer said in Sep 2026 that customer operations maps each global process digitally and checks value before deciding to expand, revise or stop an AI use case.
- Deployment model
- Unilever runs AI through a central Digital & Technology function under Chief Digital & Technology Officer Sam Kini, who reports to Chief Supply Chain and Operations Officer Willem Uijen (2025 Form 20-F), with CIO Reema Jain responsible for enterprise architecture and digital infrastructure (Feb 2026), and chief digital and information officers inside the business groups, such as Morgan Vawter for Personal Care, Home Care and Foods. Delivery sits in the functions: R&D builds its own AI tools, supply chain runs digital twins and forecasting with partners such as Accenture and Kinaxis, customer operations rolls out global agents with Microsoft, Capgemini and Genpact, and brand teams run in-house AI content studios. Unilever lets employees build small agents in a sandboxed environment, while enterprise agents follow a global rollout plan (Sep 2026).
- Model providers
- Google Cloud said in Feb 2026 that it is deploying advanced models such as Gemini for Unilever through Vertex AI, and CIO Dive reported that Unilever also plans to partner with large language model providers. Unilever customer operations staff have been trained to build regression models in Microsoft Copilot Studio (Aug 2026), and Unilever's R&D uses Microsoft Azure Quantum Elements for chemistry simulation (Jun 2024). Unilever gained access to Accenture's AI Navigator platform and a switchboard service for customizing and switching between large language models in Dec 2023. Unilever has not named the models behind its Beauty AI Studio, which it launched with an unnamed technology provider, or its R&D Assistant.
- Data and infrastructure
- Unilever completed an 18-month migration to Microsoft Azure, guided by Accenture, in Apr 2023, and Steve McCrystal, then Chief Enterprise Technology Officer, said in Feb 2025 that Unilever is 100% cloud-based. Unilever's enterprise data platform held 8 petabytes fed by 25,000 data pipelines a day, its network carried about 240 terabytes a week across more than 3 billion transactions, and its technology estate ran 1,500 applications (Feb 2025). Under the Feb 2026 Google Cloud partnership, Unilever is moving its integrated data and cloud platform and key enterprise applications to Google Cloud to build an enterprise-wide AI backbone and agentic workflows. SAP is Unilever's system of record for orders, Kinaxis runs global planning, and planners train on Databricks and Azure data tools.
Platforms and internal tools
Google Cloud data and AI backboneAnnounced Feb 17, 2026
Google Cloud data and AI backbone is the five-year Unilever and Google Cloud program to move Unilever's integrated data and cloud platform and key enterprise applications onto Google Cloud, build on Vertex AI, and develop brand discovery, measurement, AI-augmented marketing and agentic workflows.
- Models
- Google Gemini via Vertex AI
Brand DNAiBy Oct 2025
Brand DNAi is Unilever's global AI brand governance framework and brand-safe training repository, which limits AI models to a pool of approved brand voices, values, strategies and visual identities; Unilever's 2025 annual report says Beauty AI Studio is underpinned by it.
Beauty AI Studio, Personal Care AI Studio and Sketch Pro2025
Beauty AI Studio covers Beauty & Wellbeing content from ideation to personalisation for Dove, Vaseline, Pond's Skin Institute and Clear; Personal Care AI Studio was live in four markets in Mar 2026; Sketch Pro is Unilever's set of in-house AI design studios inside Home Care brand teams.
R&D Assistant and DataLabDataLab Nov 2023; R&D Assistant by Apr 2026
R&D Assistant is Unilever's own AI agent that connects more than 150,000 scientific documents from over a century of research so scientists can query them in their own languages. DataLab is Unilever's virtual global laboratory for models, simulations and analytics, used with Microsoft Azure Quantum Elements.
Forecast Engine Utility
Forecast Engine Utility combines machine learning and data science to produce a 104-week forecast every week across more than 5 million product and customer combinations in 40 operating markets (Aug 2026). A related regression model runs across about €45 billion of Unilever's turnover.
Nova and the customer operations agentsTested by Aug 2026
Nova is a Unilever agent that reads customer order emails, enters the orders into SAP and validates orders during claims. Unilever was testing Nova in Aug 2026 and was scaling it to North America in Sep 2026, after its customer operations function had rolled out eight agents worldwide.
Horizon3 LabsNov 2023
Horizon3 Labs is Unilever's AI research and development hub in Toronto, which pairs academic researchers with industry experts and technology partners; Unilever's 100+ Accelerator team planned in May 2025 to work more closely with Horizon3 Labs on AI pilots in its operations.
Vendors and partners
Google CloudFeb 2026
Data and cloud platform, Vertex AI, Gemini models, agentic commerce · Google Cloud and Unilever announced a five-year partnership on Feb 17, 2026; no deal value was published.
MicrosoftAzure migration completed Apr 2023
Azure cloud, Azure Quantum Elements in R&D, partner on global agent rollouts · Unilever's customer operations names Microsoft alongside Capgemini and Genpact as partners for enterprise agents such as Nova (Aug 2026).
AccentureDec 2023 (generative AI)
Generative AI partner; guided the Azure migration; scaling factory digital twins · Accenture's generative AI partnership with Unilever was expanded in Sep 2024 to simplify Unilever's digital foundation, and the two companies are building more than 40 digital twins across Unilever factories from Jun 2026.
Kinaxis
Global planning software · Graham Sommer said some Kinaxis-based planning processes run about 90% without manual changes (Sep 2026).
Genpact and Capgemini
Customer operations process design and delivery · Partner staff make up about half of Unilever's 6,000-person customer operations team (Sep 2026).
H2Ok InnovationsPilot reported May 2025
AI and sensor system for factory cleaning, through the 100+ Accelerator
CauzzyAISep 2026
Enterprise AI platform for Unilever Prestige, Unilever's prestige beauty unit · CauzzyAI announced the partnership in a Business Wire release; no financial terms were published.
Leadership
AI, data and platform leaders
Executives
AI org map
Central digital and technology function under a Chief Digital & Technology Officer who reports to the supply chain and operations chief; AI delivery in R&D, supply chain, customer operations and brand teams
Unilever's AI agenda is led by Chief Digital & Technology Officer Sam Kini, Global CIO until Jan 2026, who reports to Chief Supply Chain and Operations Officer Willem Uijen according to Unilever's 2025 Form 20-F. CIO Reema Jain reports to Kini, and business-group digital chiefs such as Morgan Vawter lead digital work inside the business groups. Neither Kini nor Jain sits on the Unilever Leadership Executive, the CEO's top team, and Chief R&D Officer Richard Slater's organisation runs Unilever's digital science.
Unilever does not publish the size of its AI or data science workforce. Unilever's AI work sits in the central Digital & Technology function under Sam Kini, in chief digital and information officer teams inside the business groups, in R&D's digital science and Horizon3 Labs in Toronto, and in supply chain and customer operations, where a 6,000-person team split between Unilever staff and partners runs forecasting models and agents. Unilever had trained more than 40,000 employees to use AI by Jun 2026, and about 2,000 were training to build their own AI agents in Sep 2026.
Reporting lines
Chief Executive Officer
Supply Chain and Operations
Unilever Leadership Executive member since Jan 2025; owns cyber security risk
Digital & Technology
Sam Kini, CDTO
Unilever's 2025 Form 20-F says the CDTO and the CISO report to the Chief Supply Chain and Operations Officer
Information and Technology Foundations
Reema Jain, CIO
Enterprise architecture and digital infrastructure (Feb 2026)
Business Group Digital & Information
Morgan Vawter, CDIO, Personal Care, Home Care and Foods
Reporting line not published
Customer Operations
Inferred from Unilever presenting customer operations as part of its supply chain (Aug 2026); 6,000 people
Research & Development
Richard Slater, Chief R&D Officer
Unilever Leadership Executive member since Apr 2019
Unilever's 2025 Form 20-F lists the Unilever Leadership Executive with no digital, technology or AI officer on it, and says the Chief Digital & Technology Officer and the Chief Information Security Officer report to the Chief Supply Chain and Operations Officer. Unilever has not published how the business-group digital chiefs report. Steve McCrystal, Unilever's Chief Enterprise Technology Officer for nearly seven years, joined Diageo as Chief Digital Information Officer in Jan 2026, and Andy Hill, who spoke for Unilever as Chief Data Officer in Sep 2024, joined Sainsbury's as Chief Data Officer in early 2025. Esi Eggleston Bracey, the Chief Growth and Marketing Officer who spoke for Unilever's AI content work, left the role on Jan 31, 2026, and her successor as Chief Marketing Officer is not on the Unilever Leadership Executive (2025 Form 20-F).
Named people
Use cases
Unilever has implemented more than 500 AI projects over a decade, more than 330 of them active in Sep 2024, across R&D, marketing content, factories, demand forecasting, customer operations, HR and procurement. Unilever credits AI with business returns only at single sites and tools: premium ingredient savings at its Cu Chi, Vietnam factory, cleaning costs on two Poznan, Poland lines and product image costs. Unilever reports no company-wide AI return and does not credit its €800 million productivity programme, completed in the first half of 2026, to AI.
| Use case | Area | Status | ROI |
|---|---|---|---|
| AI digital twins across the factory network Unilever runs digital twins that simulate production and recommend settings at sites including Raeford in the US and Haldia and Gandhidham in India, and is adding more than 40 AI-enabled twins with Accenture over 18 months from Jun 2026. Unilever and Accenture said in Jun 2026 that the Raeford twin for deodorant sticks predicts 95% of process flow restrictions and has delivered 20% less waste and a 10% uplift in capacity. | Manufacturing | In production | |
| Hefei smart factory Unilever's Hefei, China factory for Beauty & Wellbeing, Personal Care and Home Care uses AI and a digital twin for process optimisation and predictive maintenance alongside 3D-printed parts and line automation. Unilever says these tools together delivered an 8% rise in overall equipment effectiveness and up to 20% less wastage (Nov 2025). | Manufacturing | In production | |
| AI ingredient dosing in Cu Chi An intelligent mixer driven by an AI digital twin doses raw materials for liquid laundry products such as OMO at Unilever's Cu Chi, Vietnam factory to prevent overuse. | Manufacturing | In production | 1% to 2% saved on premium ingredients, credited to an AI digital twin mixer (Jun 2026)Source |
| Mayonnaise line digital twin in Poznan A digital twin at Unilever's Poznan, Poland foods factory steadies viscosity in mayonnaise production for brands such as Knorr and Hellmann's; Unilever and Accenture said in Jun 2026 that it cut minor stoppages by up to 20% and waste by nearly 30%, without saying the twin uses AI. | Manufacturing | In production | |
| Clean-in-place optimisation H2Ok Innovations' AI and sensor system sets detergent, temperature and duration for cleaning two process lines at Unilever's Poznan, Poland nutrition factory, cutting utility use 10% and cleaning times 20%; Unilever called it a pilot and planned a rollout to another 35 sites by 2026. | Manufacturing | Pilot | About €100,000 a year saved on two Poznan lines, credited to the AI and sensor system (May 2025) |
| Demand forecasting Unilever's Forecast Engine Utility produces a weekly 104-week forecast for more than 5 million product and customer combinations in 40 markets, and a machine learning regression model runs across about €45 billion of turnover; forecasting runs about 60% without manual changes. | Supply chain | In production | |
| Nova order and claims agent Nova reads customer order emails, enters orders into SAP and validates orders during claims; Unilever was testing Nova in Aug 2026 and scaling it to North America in Sep 2026, after rolling out eight other customer operations agents worldwide. | Customer operations | Pilot | |
| Retailer customer connectivity Unilever's AI customer connectivity model shares forecast and sales data with retailers to plan replenishment; it began as a pilot with Walmart Mexico in 2022 and was being extended to 30 key customers in Aug 2024. | Customer operations | In production | |
| Digital product twins for content Unilever keeps a digital twin of each product, with all labels, packaging and language formats in one file, and uses it with AI tools to produce product images for TV and digital commerce. | Marketing | In production | Product image costs down 50%, credited to 3D digital twins and AI (Mar 2025)Source |
| Beauty AI Studio and Personal Care AI Studio Unilever's AI content studios produce and personalise marketing assets for Beauty & Wellbeing and Personal Care brands under Brand DNAi governance, and Sketch Pro studios do the same for Home Care brand teams. | Marketing | In production | |
| Consumer insight and design semiotics Unilever Beauty & Wellbeing analyses more than 1,000 external data sources a month for consumer trends, Unilever used AI to test responses to Dove's Change the Compliment messaging, and works with an agency that uses AI-driven semiotics in Personal Care design (Oct 2026). | Marketing | In production | |
| Agentic commerce and AI discoverability Unilever and Google Cloud are building capabilities for brand discovery, conversion and measurement as shopping moves to AI assistants, and Unilever tracks how its brands appear in large language models. | Commerce | Announced | |
| R&D Assistant and AI-led formulation Beauty & Wellbeing scientists use AI over Unilever's ingredient, formulation, sensory and claims databases and query 150,000 scientific documents through the R&D Assistant; Unilever credits AI analysis of microbiome data with developing Pond's Cera-Hyamino technology. | R&D | In production | |
| Virtual cohorts and in-silico experiments Unilever builds AI virtual cohorts from its microbiome datasets to test formulas, concepts and claims on about 2,500 simulated subjects at once, alongside real-world testing. | R&D | In production | |
| Recipe Intelligence for chefs Unilever Food Solutions' Recipe Intelligence gives restaurant chefs recipes, techniques and preparation steps drawn from more than 30,000 recipes by its 250 in-house chefs. | Foods | In production | |
| Merchandiser recommendations in Brazil More than 3,600 Unilever merchandisers across 6,000 stores in Brazil use image recognition and sales data to choose which stores to visit and which shelf gaps to fix first. | Sales | In production | |
| Digital ordering for small retailers Unilever supports small-scale retailers, mainly in India and Indonesia, through AI-powered digital ordering platforms, alongside financial services and training. | Sales | In production | |
| HR chatbots, supplier onboarding and procurement Unilever made chatbots the first point of contact for most HR matters in 2025 and uses AI-enabled workflows for supplier onboarding and sourcing decisions in its global buying operations. | Back office | In production | |
| Employee-built agents Nearly 2,000 Unilever staff are training to build their own utility agents in a secure sandbox, governed more loosely than enterprise agents such as Nova. | Workforce | Pilot |
Adoption
Unilever spread AI in stages: AI in its systems for more than a decade (Steve McCrystal, Feb 2025), an AI assurance review from 2019, a full move to Microsoft Azure in 2023, and training for the whole workforce from 2024 under a go wide and go deep plan. Unilever trained about 16,000 employees in generative AI by Sep 2024, 23,000 by the end of 2024 and more than 40,000 by Jun 2026, out of 96,092 employees at Dec 31, 2025. Unilever has published regular use only for one business group: more than 75% of office-based Personal Care employees used personal AI productivity tools regularly in Nov 2025. In supply chain, the DigiOps program gives employees about 150 hours of on-the-job training over nine months, and Graham Sommer said markets with more DigiOps learners show higher adoption (Sep 2026).
By the numbers
Rollout, first pilot to today
2019
Unilever created its AI assurance process to review AI systems for ethical and performance risks before deployment.
2022
Unilever piloted its AI customer connectivity model with Walmart Mexico, linking retailer sales data to Unilever's forecasting and replenishment.
Apr 2023
Unilever completed an 18-month migration to Microsoft Azure, guided by Accenture.
Nov 2023
Unilever opened Horizon3 Labs, an AI research hub in Toronto, and launched DataLab, its virtual R&D laboratory.
Dec 2023
Unilever and Accenture agreed a generative AI partnership giving Unilever access to Accenture's AI Navigator platform and a switchboard for moving between language models.
Jun 2024
Unilever described its use of Microsoft Azure Quantum Elements with DataLab to run chemistry and formulation simulations.
Sep 2024
Unilever said it had more than 500 AI projects, more than 330 active, had put about 150 through AI assurance review and had trained about 16,000 employees in generative AI that year.
Feb 2025
Steve McCrystal said Unilever had trained 23,000 employees in AI by the end of 2024 and was 100% cloud-based.
Mar 2025
Unilever reported that digital product twins and AI tools were producing product images at half the previous cost, an approach Esi Eggleston Bracey, then Chief Growth and Marketing Officer, presented at an NVIDIA conference.
Oct 2025
Unilever described Beauty AI Studio, Sketch Pro, Recipe Intelligence and its Brand DNAi governance framework as the base of AI content across its brands.
Nov 2025
Unilever said more than 75% of office-based Personal Care employees had become regular users of personal AI productivity tools.
Feb 2026
Unilever and Google Cloud announced a five-year partnership to move Unilever's data and cloud platform to Google Cloud and build on Vertex AI.
Jun 2026
Sam Kini said Unilever had trained more than 40,000 employees to use AI, and Unilever began adding more than 40 AI digital twins with Accenture.
Sep 2026
Unilever customer operations had rolled out eight AI agents worldwide and was training nearly 2,000 people to build their own agents.
Economics & controls
Spend and ROI
Technology budget
Unilever does not publish a technology or AI budget, and neither Unilever nor Google Cloud gave a value for their Feb 2026 partnership. Unilever's 2025 restructuring costs were €599 million, 1.2% of turnover, concentrated in supply chain, productivity initiatives and functional transformation, and Unilever said in Jul 2026 that most of its €131 million of first-half restructuring costs related to one-off technology and supply chain projects. Unilever announced in May 2026 that it will invest $270 million over the long term, including $50 million of capital expenditure, in a digital-first, AI-powered R&D centre in New Haven, Connecticut, for Beauty, Wellbeing and Personal Care, which it expects to be fully operational by spring 2029.
Reported ROI
Unilever reports no company-wide return on AI. The returns Unilever credits to AI are site and tool level: 1% to 2% savings on premium ingredients at Cu Chi from an AI digital twin mixer, about €100,000 a year from AI-guided cleaning on two Poznan lines, and product images at half the previous cost with 3D digital twins and AI. Unilever credits Hefei's 8% rise in overall equipment effectiveness to AI together with 3D printing and line automation, and does not say whether the Poznan mayonnaise twin, which cut waste by nearly 30%, uses AI. Unilever's €800 million productivity programme, launched in 2024 and completed ahead of plan in the first half of 2026, cut overheads by 50 basis points in 2025, but Unilever does not credit those savings to AI.
Productivity markers
Unilever Beauty & Wellbeing says AI lets its teams analyse consumer insights 60% faster, cut formulation rounds from five or six to one or two, and generate claims 75% faster (Apr 2026), and Beauty AI Studio produces assets up to 30% faster (Oct 2025). Sam Kini said in Jun 2026 that AI-powered digital twins are helping boost production capacity 10% at Unilever's Raeford, North Carolina personal care factory, a capacity measure for which Unilever gives no cost or revenue equivalent. Fernando Fernandez said in Jun 2026 that Unilever's innovation timelines have shortened from two to three years to nine to 12 months as 80% of laboratory work became digital. Graham Sommer said in Sep 2026 that forecasting runs about 60% without manual changes and some Kinaxis planning processes about 90%.
Headcount
Unilever had 96,092 employees at Dec 31, 2025, down from 120,040 a year earlier, mainly because of the Ice Cream demerger in Dec 2025, and its employee turnover rate rose to 17.2% from 14.5%, which Unilever attributes mainly to the productivity programme. Unilever's 2025 annual report says the company shifted from time-intensive, people-centred processes to technology and AI solutions in back-end operations in 2025, but Unilever has not tied a headcount figure to AI.
Portfolio changes
Unilever demerged its Ice Cream business on Dec 6, 2025 and agreed in Mar 2026 to combine its Foods business with McCormick, with completion expected by mid-2027 at the latest. Unilever Food Solutions' Recipe Intelligence and the Poznan nutrition factory's AI cleaning system sit in the Foods business that is due to leave Unilever.
Governance and risk
AI assurance review
Unilever created an AI assurance process in 2019 in which a cross-functional team reviews AI systems for ethical and performance risks. Unilever told CIO Dive in Sep 2024 that about 150 prioritised projects had gone through the review and that around half needed fixes for performance, unwanted bias or transparency. Steve McCrystal, then Chief Enterprise Technology Officer, said in Feb 2025 that a Responsible AI framework governs how Unilever develops, deploys and uses AI.
Policies and principal risks
Unilever's 2025 Form 20-F says three company policies were updated for AI risks in intellectual property, data privacy and marketing, and that generative AI raises the risk of misuse, loss of trust and legal liability. Unilever lists AI inside its business transformation risk, warning that design choices in its programme to modernise digital foundations could embed complexity, and says attackers use AI for phishing and deepfake social engineering.
Brand safety
Unilever's Brand DNAi framework limits AI content models to approved brand voices, values, strategies and visual identities, and Unilever's 2025 annual report calls it the company's global AI brand governance framework.
Controls on agents
Reema Jain said in May 2026 that Unilever's foundation for agents rests on governance, observability and interoperability, with each agent given clear business ownership, explicit objectives and defined guardrails. Graham Sommer said in Sep 2026 that enterprise agents such as Nova follow a global rollout plan with governance built in, while employees experiment with small agents in a secure sandbox that Unilever does not try to control centrally.
AI literacy and the EU AI Act
Unilever tied its 2024 generative AI training to Article 4 of the EU AI Act on AI literacy, with courses that range from basic practice to university-level data science, and prepared its use cases for the Act's risk-management and bias-review provisions (Sep 2024).
Human oversight in R&D
Unilever says scientists oversee its AI virtual cohorts and that the cohorts complement real-world testing without replacing it (Apr 2026).
Recent signals
Oct 28, 2026
Unilever's third-quarter trading statement upgraded its 2026 outlook, named rapid advances in generative AI among its risk factors, and set an Investor Event in New York for Nov 4, 2026.
Sep 28, 2026
Graham Sommer said Unilever customer operations has rolled out eight AI agents worldwide, is scaling the Nova order agent to North America and is training nearly 2,000 people to build their own agents.
Sep 15, 2026
Jason Harcup said Unilever Beauty & Wellbeing plans further investment in predictive science, digital twins, agentic AI and personalised consumer experiences.
Sep 2026
CauzzyAI announced a partnership to bring its enterprise AI platform into Unilever Prestige, Unilever's prestige beauty unit.
Aug 20, 2026
Vicky Cuthbert said Unilever's Forecast Engine Utility forecasts 104 weeks ahead every week across 5 million product and customer combinations, and that the Google Cloud partnership will support agentic workflows across business processes.
Jun 16, 2026
Sam Kini said Unilever had trained more than 40,000 employees to use AI and called becoming AI-first one of three foundational pillars of Unilever's transformation.
Jun 16, 2026
Unilever and Accenture began scaling AI-enabled digital twins across Unilever's factories, with more than 40 to be built over 18 months.
Jun 5, 2026
Fernando Fernandez said 80% of Unilever's laboratory work is now digital and that on Amazon Unilever tracks how more than 50 of its brands appear in large language models in 20 geographies.
May 21, 2026
Reema Jain said Unilever deploys AI agents with clear business ownership, explicit objectives and defined guardrails, and sees agent applications in supply chain, procurement and customer operations.
May 2026
Unilever announced a $270 million digital-first, AI-powered R&D centre in New Haven, Connecticut, for its Beauty, Wellbeing and Personal Care brands, due to be fully operational by spring 2029.
Apr 29, 2026
Unilever Beauty & Wellbeing described its R&D Assistant agent over 150,000 scientific documents and AI virtual cohorts of about 2,500 simulated subjects.
Mar 24, 2026
Steve McCrystal, Unilever's Chief Enterprise Technology Officer for nearly seven years, was reported to have joined Diageo as Chief Digital Information Officer in Jan 2026.
Mar 12, 2026
Unilever's 2025 Form 20-F said its Chief Digital & Technology Officer reports to the Chief Supply Chain and Operations Officer, that three policies were updated for AI risks, and that chatbots became the first point of contact for most HR matters in 2025.
Feb 27, 2026
Unilever named Reema Jain CIO, reporting to Sam Kini, who had served as CIO until Jan 2026 and became Chief Digital & Technology Officer.
Feb 17, 2026
Unilever and Google Cloud announced a five-year partnership covering Unilever's data and cloud platform, Vertex AI, Gemini models and agentic commerce.
Quick answers
- How does Unilever deploy AI?
- Unilever credits AI with returns in its factories and marketing, such as 1% to 2% ingredient savings at its Cu Chi plant and product images at half the cost, but reports no company-wide AI figure. Unilever has implemented more than 500 AI projects and in Feb 2026 announced a five-year deal to move its data and cloud platform to Google Cloud and build on Vertex AI and Gemini. Chief Digital & Technology Officer Sam Kini said in Jun 2026 that more than 40,000 employees had been trained to use AI.
- What does Unilever use AI for?
- Unilever has implemented more than 500 AI projects over a decade, more than 330 of them active in Sep 2024, across R&D, marketing content, factories, demand forecasting, customer operations, HR and procurement. Unilever credits AI with business returns only at single sites and tools: premium ingredient savings at its Cu Chi, Vietnam factory, cleaning costs on two Poznan, Poland lines and product image costs. Unilever reports no company-wide AI return and does not credit its €800 million productivity programme, completed in the first half of 2026, to AI.
- Who leads AI at Unilever?
- AI at Unilever is led by Sam Kini (Chief Digital & Technology Officer); Reema Jain (Chief Information Officer); Meenakshi Burra (Chief Data Officer). Ownership of AI on the org chart: CDTO Sam Kini, reporting to supply chain and operations chief Willem Uijen.
- What AI platform does Unilever use?
- Unilever builds AI into its own tools, such as Brand DNAi for brand-safe content and an R&D Assistant agent over 150,000 scientific documents, and in Feb 2026 chose Google Cloud's Vertex AI as the base of an enterprise data and AI backbone.
- Which AI models does Unilever use?
- Google Cloud said in Feb 2026 that it is deploying Gemini for Unilever; Unilever teams also build with Microsoft Copilot Studio, and Unilever has named no other model provider.
- How many Unilever employees use AI?
- Unilever had trained more than 40,000 of its roughly 96,000 employees to use AI by Jun 2026, and more than 75% of office-based Personal Care staff used personal AI productivity tools regularly in Nov 2025.
- How far has AI spread at Unilever?
- Unilever credits AI with business returns in two functions: in manufacturing, an AI digital twin mixer saves 1% to 2% of premium ingredients at its Cu Chi, Vietnam factory (Jun 2026) and an AI cleaning system saves about €100,000 a year on two Poznan, Poland lines (May 2025); in marketing, product images made with AI-driven product twins cost 50% less (Mar 2025). Unilever gives no firm-level AI figure and does not credit its €800 million productivity programme to AI. Unilever has not reported how many employees use its AI tools: Chief Digital & Technology Officer Sam Kini, who owns the AI-first agenda, said in Jun 2026 that AI-powered digital twins are cutting energy use and optimising ingredients in Unilever's factories and that more than 40,000 employees had been trained to use AI, which counts training rather than use, and Unilever's only regular-use figure covers office-based Personal Care staff. Unilever is therefore at Scaling, at its low edge because its returns are site and tool level and its Google Cloud data and AI backbone is still being built; Unilever falls short of Embedded because it credits no firm-level cost, revenue or margin figure to AI and reports no weekly AI use by most of its 96,092 employees.
Open questions
- Which general-purpose AI assistant do Unilever employees use, and how many of the 40,000-plus trained staff use AI tools every week across the company?
- Does anyone on the Unilever Leadership Executive own AI results, given that Sam Kini reports to Willem Uijen and sits below the executive team?
- Who leads data and enterprise technology at Unilever now: DataIQ named Meenakshi Burra Chief Data Officer in 2026, and Unilever has not said who took over Steve McCrystal's remit after he left for Diageo?
- How much is Unilever spending on the Google Cloud partnership, and which workloads stay on Microsoft Azure?
- Which model providers besides Google's Gemini does Unilever use, and does it still route between models through Accenture's switchboard?
- How many of Unilever's more than 500 AI projects are in production today, given that the last split, more than 330 active, dates from Sep 2024?
- What business return do Unilever's deployed tools without published figures deliver, such as the Forecast Engine Utility, Beauty AI Studio, the R&D Assistant and the HR chatbots, and how much of the Hefei factory's gains comes from AI alone?
- What is the 10% capacity gain from Unilever's Raeford digital twin worth in cost saved or revenue, and do the other factory twins report returns in currency?
- How much of Unilever's €800 million productivity programme came from AI and automation, and will Unilever ever report a company-wide AI return?
- What happens to the AI tools in Unilever Foods, such as Recipe Intelligence and the Poznan cleaning system, when the business combines with McCormick?
- Will Unilever's Nov 4, 2026 Investor Event in New York add AI targets or figures?
Sources
Unilever
- How Unilever is building an AI-first enterprise at scale, Sam Kini (Jun 16, 2026)
- Unilever scales digital twins across global manufacturing network with Accenture (Jun 16, 2026)
- Unilever unveils plans for $270 million US-based global innovation centre (May 2026)
- Unilever reinvents product shoots with digital twins and AI (Mar 2025)
- Inside Unilever's AI-powered supply chain, Vicky Cuthbert (Aug 20, 2026)
- How AI is transforming innovation in Unilever Beauty & Wellbeing (Apr 29, 2026)
- How design is driving desire across Unilever's Personal Care portfolio (Oct 27, 2026)
- What's behind Unilever's 2026 first half results (Jul 31, 2026)
- How AI is transforming Unilever's Personal Care business (Nov 14, 2025)
- How AI is helping drive Desire at Scale across Unilever (Oct 3, 2025)
- How we're future-proofing talent across our supply chain (Sep 1, 2025)
- Unilever's 100+ Accelerator partnership and the H2Ok pilot (May 16, 2025)
- How Unilever's digital transformation is driving operational excellence, Steve McCrystal (Feb 11, 2025)
- Utilising AI to redefine the future of customer connectivity (Aug 5, 2024)
- Future R&D: how Unilever is transforming innovation with Microsoft (Jun 24, 2024)
Filings and results
Partners and events
- Google Cloud: Unilever and Google Cloud partner on consumer goods technologies (Feb 17, 2026)
- London Tech Week: Sam Kini speaker profile (Jun 2026)
- London Tech Week: Meenakshi Burra speaker profile
- DataIQ 100 Brands 2026: Meenakshi Burra, Chief Data Officer and CDIO, Unilever
- Business Wire: CauzzyAI and Unilever Prestige partner on AI (Sep 2026)
News
- Consumer Goods Technology: How Unilever scales AI, from predictive models to agentic workflows (Sep 28, 2026)
- Consumer Goods Technology: How CGs like Unilever establish foundations for scalable AI (Aug 31, 2026)
- Consumer Goods Technology: How Unilever's FIFA product plan came to life with AI and digital twins (Aug 14, 2026)
- Consumer Goods Technology: Unilever to scale agentic twin development (Jun 16, 2026)
- Consumer Goods Technology: Unilever reengineers R&D and marketing for the AI age (Jun 5, 2026)
- Consumer Goods Technology: Unilever, Newell and Levi's on agentic workforces (May 21, 2026)
- Consumer Goods Technology: Unilever names Reema Jain CIO (Mar 3, 2026)
- CIO Dive: Unilever taps Reema Jain as CIO (Feb 27, 2026)
- Storyboard18: Diageo appoints Steve McCrystal as Chief Digital Information Officer (Mar 24, 2026)
- CDO Magazine: Sainsbury's appoints Andrew Hill as Chief Data Officer (Mar 12, 2025)
- Consumer Goods Technology: Unilever optimizes product shoots with digital twins and AI (Mar 21, 2025)
- Premium Beauty News: Unilever leans into AI for a new era of beauty innovation (Sep 15, 2026)
- CIO Dive: Unilever targets agentic AI with Google Cloud deal (Feb 17, 2026)
- CIO Dive: Unilever readies use cases and adoption plans for EU AI Act provisions (Sep 23, 2024)
- CIO Dive: Accenture and Unilever partner on generative AI (Dec 6, 2023)
- CIO.inc: Unilever's AI push from shop floor to culture core, part 1 (Aug 2025)
- CIO.inc: Unilever's AI push from shop floor to culture core, part 2 (Aug 12, 2025)
Data confidence
- Confirmed
- Stated in a primary or reputable public source.
- Estimated
- Derived or approximated from public reporting.
- Unconfirmed
- Reported, but not yet corroborated.
Unilever profile · Enterprise AI diffusion report by Imagine · facts checked against their sources on Nov 1, 2026
