The beauty industry makes artificial intelligence a partner in product creation | artificial intelligence

The tool is fed with thousands of data regarding formulas, ingredients, fragrance history, and consumer acceptance rates. “The launch was a media success rather than a sales success,” recalls Gustavo Diamant, R&D director at Grupo Boticário. “It was a learning experience for us. It was the beginning of using AI. We pursued AI without listening to the consumer.”

Six years later, another fragrance, Floratta Red Passion, appeared on the market with the help of artificial intelligence. The difference is that the technology was used to achieve a fragrance blend developed by analyzing the comments of more than 50 thousand consumers about women’s fragrances, which were drawn through the specialized fragrance platform, using artificial intelligence technology. “Over the years, we’ve improved the use of the tool, which now exists on multiple fronts, from creating new products to mapping launch opportunities,” says Diamant.

With the help of Quantic Model, a predictive model for product development, Boticario was able to increase researcher productivity by 57%. “In the past, to reach the ideal formula, 70 combinations were developed. Now, artificial intelligence, armed with information about the desired product, generates three formulations of the required quality, through which the researcher reaches the final formula,” reveals Diamant. According to him, the smart lipstick, which targets people with limited mobility or visual impairment, is an example of a product created using artificial intelligence. “The project, in its fifth phase, has a prototype that uses advanced sensors and algorithms to differentiate between facial and lip skin, and apply lipstick accurately and uniformly,” says the CEO.

According to the “Global Artificial Intelligence Survey”, conducted by consulting firm McKinsey, 72% of organizations from various sectors adopted some artificial intelligence models in the past year. In 2018, it did not exceed 50%. The beauty industry is no different. According to data from the Zipdo Education Report 2025, 45% of skincare brands are integrating AI into their R&D processes or formulations. More than 30% of new skincare products launched in 2023 included AI-based ingredients or processes. “Anyone who doesn’t use it is already considered obsolete,” says Ariadne Moraes, director of innovation and regulatory affairs at the Brazilian Association of the Toiletries, Perfumes and Cosmetics Industry (ABIHPEC). “AI has moved beyond the utility phase in the beauty sector, as it has become a co-creative tool in the development of formulas, new active ingredients, and refining formulations, among other fronts.”

A supplier of fragrances and scents to large industries such as Unilever, Boticário and Natura, German company Symrise maintains a global team focused exclusively on developing AI solutions. “By sharing data, technology helps us create combinations that are closer to user requirements,” says Lais Aragon, Senior Director of Consumer and Market Insights in Latin America. “AI is not independent, it enters the process as an ally to help us find more effective ways to deliver the best solutions to the consumer.” According to her, the challenge lies in combining artificial intelligence with the specific, almost literal knowledge of the perfumer. “We are still learning,” he says.

As hyper-personalization gains more and more evidence, using AI to create niche-focused products is becoming a strategy embraced by the beauty industry across the board. “There is a lot of talk about personalization, however, what is offered is still mainly recommendations, not products,” says Marcela Andrade, project manager at Gouvêa Consulting. “It would be difficult to offer something completely customized, but what we are looking for is to develop something very close to ideal with the help of artificial intelligence.”

In the view of Romulo Zamberlan, director of advanced research at Natura, the most challenging thing from an AI point of view is not the technology, but the speed of transformation required by the tool, which requires constant adaptation to the mental model. “When we talk about applying AI in research, biodiversity in the Amazon, omics data (biological information generated by high-throughput technologies, which allows large-scale analysis of a range of molecules), bioinformatics and ingredient safety, we are talking about well-calibrated models, robust databases, as well as multidisciplinary teams able to interpret results with scientific rigor.” Another sensitive point, from the executive’s point of view, is ensuring that the acceleration brought by AI does not exceed regulatory and safety requirements, while maintaining the focus on human well-being and respect for nature.

In 2024, Natura was highlighted at IFSCC, the largest scientific conference in the global cosmetics industry, with an article on a predictive model for the safety and efficacy of bioactive ingredients using artificial intelligence. “This technology expands our ability to analyze thousands of ingredient combinations, predict biological effects, improve formulas and fragrances, and study skin, hair and microbiome properties more precisely,” says Zamberlan. “Next steps include deepening the integration of artificial intelligence, metrology and bioinformatics, expanding the ability to read biodiversity and translating this into renewable cosmetic technologies.”

For Fabio Cahen, Basf’s personal care business manager in South America, there are opportunities in many areas using AI in the personal care industry. “The ability to analyze large amounts of data gives us more flexibility to identify new ingredients and trends and develop new formulations,” he says. “Enabling rapid and efficient prototyping is a promising frontier, as it reduces the time between product development and launch, as well as promoting more sustainable solutions.”