The story of artificial intelligence has long been dominated by chips, data centers, and the companies building the underlying models. But a quieter revolution is taking place far from Silicon Valley — in the shampoo aisle, the cookie jar, and the soap section of your local supermarket. The world's largest makers of everyday consumer goods are now embedding AI into the very heart of their research and development processes, reshaping how products are conceived, formulated, and brought to market.
Procter & Gamble, the sprawling consumer conglomerate behind brands like Pantene, Tide, and Gillette, offers one of the most striking examples. The company reports that it used AI to screen tens of thousands of peptides in developing a new formula for a Pantene product. This was made possible by an internal database containing more than 8,500 formulations, which allowed algorithms to predict how a mixture would feel on skin or hair before any physical mixing took place. The traditional method — iterative lab tests, trial and error, and extensive sensory panels — can take months. With AI, many of these steps are collapsed into computational simulations, pushing candidate products toward consumer trials at a much faster pace.
From R&D to the retail shelf
The implications are profound. In the consumer goods industry, the cost of experimentation has historically been measured in months of lab work and test batches. AI turns this into a search problem over known ingredients, dramatically reducing the time from concept to prototype. P&G's approach mirrors a broader trend across the sector, where established giants are racing to integrate machine learning and generative models into every stage of product creation.
Mondelez, the snacking powerhouse behind Oreo, Cadbury, and Ritz, is pursuing a similar transformation on the food side. The company has deployed an AI product-development tool that helps its teams generate dozens of new formulations quickly. According to Mondelez, the software enables developers to move between two and five times faster than conventional methods. This means more variants, faster iterations, and a constant churn of new flavors and textures hitting the shelves. For consumers, the visible result is an ever-expanding array of options — limited-edition cookies, seasonal chocolate bars, and novel snack combinations that appear and disappear with increasing frequency.
But AI is not only reshaping what goes inside the package; it is also transforming how those packages are marketed. The same generative systems that assist in formulation are being turned outward, producing personalized images, text, and video at a scale that traditional creative studios cannot match. This is where Unilever, the Anglo-Dutch consumer giant behind Dove, Hellmann's, and Ben & Jerry's, has leaned hardest.
Marketing meets machine learning
Unilever's Dove brand recently launched a cookie-scented body-care line in partnership with Crumbl, the popular cookie chain. The campaign was a showcase for AI integration across the entire value chain. From product direction — identifying the scent and texture trends that would resonate — to the selection of influencers and the generation of creative assets, AI played a central role. The company reported that the campaign generated billions of impressions and attracted a significant share of new buyers to the brand. Whether one finds the idea of cookie-scented soap appealing or absurd, the mechanics are instructive. A single AI-assisted pipeline now runs from formulation to feed, compressing what once required multiple agencies and months of coordination into a streamlined, data-driven process.
What ties these examples together is compression. In consumer goods, the traditional cost of a campaign is measured in agency hours and weeks of production. AI attacks this by enabling on-demand generation of content — dozens of variations of an ad, each tailored to a specific audience segment or platform. This mirrors the advertising ambitions on display when OpenAI pitched AI-made ads at the Cannes Lions festival, signaling that the technology is no longer just a back-office efficiency tool but a creative force in its own right.
Caution and skepticism remain
However, the claims deserve some caution. Most of the specific figures cited come from the companies themselves, and consumer giants have every incentive to present their AI programs as more advanced than they actually are. Product development still ends with human tasting panels and dermatological testing — an algorithm can predict performance, but it cannot replicate the nuanced judgment of a trained sensory expert. A formula that scores well in a computational model is not the same as one that a shopper reaches for repeatedly. The industry's own researchers have flagged that AI-generated marketing often drifts toward the generic, losing the brand-specific character that makes a campaign resonate emotionally with consumers.
Moreover, the technology is still evolving. AI models trained on historical data may struggle to anticipate truly novel consumer preferences or breakthrough ingredient combinations. They are, by nature, backward-looking, optimizing for what has worked before rather than pioneering entirely new categories. This can lead to incremental innovation rather than radical leaps, a limitation that companies must navigate carefully.
Broader enterprise adoption
Despite these caveats, the direction of travel is clear and consistent across firms that rarely agree on much. The reallocation of enterprise budgets toward AI agents and tooling has become a general feature of large companies, from Tencent's enterprise agents in China to the consumer-goods R&D described here. The packaged-goods sector is not sitting out. Investment in AI research labs, partnerships with technology vendors, and internal upskilling programs are accelerating. Companies like P&G have established dedicated digital innovation teams that work alongside traditional chemists and marketers, fostering a hybrid approach that blends domain expertise with machine intelligence.
The historical context is important too. Consumer goods giants have long been early adopters of process automation — from robotic assembly lines to supply chain optimization algorithms. AI represents the next logical step, extending automation into the creative and scientific domains. The difference is that unlike previous waves of digitization, generative AI can produce original outputs — whether a new fragrance formula or a video advertisement — rather than simply optimizing existing workflows.
For shoppers, the visible result will be mundane but significant: more product variants, faster refresh cycles, and scents and textures that arrive and vanish more quickly than they used to. The machinery behind the shelf is changing even where the products look the same. A bottle of shampoo is, increasingly, the output of a search — an algorithm's best guess at what will appeal to a particular demographic at a particular moment. The same is true for the cookie on the snack rack, the detergent on the laundry aisle, and the moisturizer on the beauty counter.
The implications extend beyond consumer choice. Shorter development cycles mean less waste — fewer unsold products sitting in warehouses — and more responsive supply chains. They also raise questions about employment: as AI takes on formulation and creative tasks, what role remains for the human chemists, flavorists, and copywriters who have traditionally driven innovation? The answer, at least for now, is that human oversight remains essential, but the nature of that oversight is shifting from doing to directing, from creating to curating. The most successful companies will be those that find the right balance between algorithmic efficiency and human intuition.
In the end, the AI story in consumer goods is one of compression and acceleration. The time from idea to shelf is shrinking, the cost of experimentation is dropping, and the boundaries between product development and marketing are blurring. As these trends converge, the shampoo aisle becomes a window into a broader transformation — one where the outputs of a model increasingly shape the objects we hold in our hands each morning.