"Complexity is not a problem to eliminate — it's a skill to master." That was the guiding thread of the keynote by [speaker to be confirmed], who presented at Tech For Industry Show 2026 (Paris Expo Porte de Versailles) the concrete AI use cases across L'Oréal's operations — from generative AI for innovation to warehouse robotics, predictive maintenance and demand forecasting.

Behind the group's 40 brands lies a little-known industrial reality: €44 billion in revenue, 7 billion products manufactured and distributed every year across 150 countries, and 22,800 operations experts. L'Oréal controls its entire value chain — design, procurement, production, distribution — which, in the AI era, delivers a decisive advantage: end-to-end data ownership, from design data to consumer data.

The industrial footprint: 37 factories worldwide, including 11 in France, producing close to consumer markets, and 160 distribution centers serving eight different channels, from mass retail to direct e-commerce, pharmacies and hair salons. The goal is simple to state and formidable to execute: the right product, in the right place, at the right time.

The catalog counts 72,000 SKUs, from personalized niche products to mass-market items produced by the millions. On the industrial side: around 1,000 packaging lines — from low-speed multi-format lines to single-format lines running at 200 strokes per minute — and 350 manufacturing vessels ranging from 50 kg to 20 tonnes.

Beauty is an innovation-driven market: around 8,000 new products per year and a 30% catalog renewal rate. That means new packaging, new formulas, sometimes new manufacturing processes — a growth engine that is also a permanent operational challenge.

A TikTok trend or an influencer praising a lipstick shade, and demand explodes with zero warning. Great commercial news, operational headache: ingredients must be secured, production ramped, delivery accelerated. Add to that the world's complexities: supply chain fragmentation, geopolitical instability, labor shortages, climate risks and increasingly fast, fragmented regulatory change.

Facing this complexity, L'Oréal's operations focus on three priorities. Quality, the historic pillar: around 100 checks for every product that reaches a consumer's hands. Agility: speed of innovation, of reaction to the unexpected, of delivery. Sustainability, with a major milestone reached in 2025: 100% of directly operated sites powered by renewable energy, under the L'Oréal for the Future program — recognized by CDP for the 10th consecutive year and an EcoVadis Platinum medal for the 4th consecutive year.

The stated approach is pragmatic: first, a robust data strategy — every new use case reveals missing data or quality gaps to fix; second, a global approach, with business-line centers of excellence that test solutions, validate ROI and then deploy them worldwide; third, the discipline of focusing on high-ROI use cases serving the company's priorities, resisting the temptation to do everything.

GenAI produces more ideas, visuals, renders and mockups, much faster, to feed the innovation pipeline. In execution, digital simulation replaces part of physical testing with virtual tests — more tests, faster, cheaper — including for regulatory compliance updates, which keep multiplying.

A globally deployed sales forecasting solution leverages the mass of available data to improve accuracy and detect weak signals — the key to reacting fast when a product goes viral.

The highly automated industrial base is equipped with sensors and IoT systems generating a continuous, exploitable data stream. AI detects process drifts in real time and gives operators, technicians and managers structured information for fast, autonomous decisions. Three use cases:

Widely deployed vision systems now reach unprecedented accuracy thanks to AI — down to detecting defects on transparent glass jars, previously very hard to identify — and characterize defects in the flow for faster corrective action.

Identifying equipment performance drift makes it possible to act before a failure impacts production.

Any abnormal water or electricity consumption is detected immediately, and connected parameters pinpoint the cause. The payoff: gains in quality, waste reduction, resource consumption and equipment yield.

On the foundation of a single WMS deployed across all group distribution centers, AI models are being tested to better orchestrate millions of order lines within tight deadlines. Robotics automates a growing share of warehouse movements, and new robot generations with embedded AI open the way to more flexible automation.

The keynote's conclusion puts humans back at the center: an augmented workforce, capable of faster, better-informed decisions, with clear-eyed honesty — some jobs will disappear, others will transform, new ones will emerge. The manufacturer's responsibility: ensuring upskilling and training, the necessary condition for a transformation that is "committed, responsible and performance-driving."

How does L'Oréal use artificial intelligence in its factories? Mainly through three use cases: vision-based quality control (in-line defect detection and characterization), predictive maintenance, and real-time detection of process drifts and abnormal water or energy consumption.

How can a supply chain absorb a viral demand spike? Through visibility: L'Oréal has deployed a global AI sales forecasting solution that detects weak market signals, making it possible to secure ingredients, produce and deliver faster when a product goes viral.

How should companies select AI use cases? L'Oréal's approach rests on three pillars: a robust data strategy, centers of excellence that test and validate ROI before global rollout, and strict focus on high-ROI use cases aligned with company priorities.

This conference was recorded at Tech For Industry Show 2026, the B2B trade show for industrial transformation (Smart Manufacturing, AI, Supply Chain) at Paris Expo Porte de Versailles. Watch the full keynote video above, and join us on May 19–20, 2027, Hall 7, for the next edition.