AI in Product Development: How L'Oréal, Nestlé and Mondelēz Are Using It

AI in product development shortens time to market: L'Oréal has used the technology in its labs for four years to predict molecular behavior and find new uses for ingredients.

How L'Oréal uses AI in formula development

The technology helps predict molecular behavior, one of the most labor-intensive stages in creating new cosmetic products. Developing a new formula has traditionally been a long, expensive process: chemists synthesize and test compounds by hand, assess ingredient compatibility, and check the stability and safety of formulations. For every product that reaches the shelf, dozens of formulas are discarded at intermediate stages.

An AI model can predict the outcome before any physical experiment begins: researchers get a list of promising hypotheses and test only those. The number of dead ends and the cost of each iteration drop sharply.

The second use case is finding new applications for existing components. L'Oréal has built up an extensive library of ingredients, and the algorithm identifies which of them can be used in new products or contexts without starting the research cycle from scratch.

The use of AI in lab R&D is part of a broader industry trend: machine learning has been used in pharmaceuticals for several years to screen molecules, and similar approaches are now spreading to consumer chemistry and cosmetics.

• Four years: how long L'Oréal has used AI in its labs
• Fabrice Megarbane: president of L'Oréal's consumer division
• AI tasks: predicting molecular properties and finding new uses for ingredients
• Companies in the trend: L'Oréal, Mondelēz, Nestlé

Why FMCG giants are betting on AI in R&D

Mondelēz and Nestlé, among the world's largest consumer goods producers, are moving in the same direction. Three companies from different sectors (cosmetics, confectionery, food) are applying AI to streamline development at the same time.

The simultaneous move by three different players is no coincidence: all face a similar structural problem. Consumers expect increasingly personalized products, retailers demand constant refreshes of the assortment, and competition from private labels and direct-to-consumer brands is intensifying. AI offers a way to do more with the same resources.

A traditional FMCG R&D cycle takes years: molecular research, formula iterations, consumer testing, safety checks, regulatory approval. AI takes over the most analytical and repetitive parts of the cycle and frees teams for tasks that require expert judgment. In a fast-changing industry, development speed becomes a competitive advantage: a company that launches a product in 18 months instead of three years claims shelf space first and tests market response first.

What this means

When L'Oréal, Nestlé and Mondelēz all report practical use of AI in product R&D, it signals a systemic shift in traditional industries. Four years of real-world use in L'Oréal's labs point to measurable results, not experimental curiosity. The industry's next question: how much exactly has AI cut development time and cost in specific product lines, and when will companies start publishing that data.

How do companies use AI in product development?

AI helps predict molecular behavior, which shortens development time, makes it possible to find new uses for existing ingredients and speeds up R&D cycles.

Which companies have already adopted AI in development?

L'Oréal has used AI in its labs for four years; Mondelēz and Nestlé are taking a similar approach to speed up development.