Beauty brands use AI across R&D and consumer tools to accelerate development and personalization
AI has moved from experimental to operational across beauty company workflows. Ingredient discovery, safety testing, regulatory compliance, and consumer-facing product recommendations now rely on AI systems to handle tasks that previously required more time and manual work.
Amorepacific, a K-beauty company, recently used AI to identify an optimal peptide for hair strengthening. The company analyzed approximately 8,000 compounds through computational simulation and molecular dynamics analysis to find Tripeptide-132, which binds most strongly to hair keratin. The research was published in the International Journal of Cosmetic Science.
The work demonstrates how AI can compress the ingredient discovery cycle. Amorepacific's team used AI to interpret complex biological interactions and optimize formulation strategies-tasks that traditionally required longer experimental cycles.
Safety testing and regulatory work accelerate
Beyond ingredient discovery, Amorepacific deployed AI for safety assessment. An automated skin irritation evaluation system trained on approximately 83,000 patch test images speeds up clinical research and safety verification across markets.
The company also uses AI to manage cosmetic compliance. Different countries enforce different labeling and ingredient regulations. AI systems trained on country-specific rules help prepare regulatory documentation faster and with fewer errors.
Consumer-facing AI requires credibility to drive adoption
On the consumer side, AI-powered skin analysis and product recommendation tools are becoming standard. Many brands now offer virtual consultations and personalized routines based on skin assessment.
Adoption varies widely, though. Consumers quickly disengage from AI that feels generic or disconnected from their actual skin needs. Quiz-based recommendations without objective skin analysis often fail to build user trust.
AI experiences that combine direct skin analysis with additional data-such as geolocation to assess UV exposure or lifestyle factors-perform significantly better. When users sense the system is grounded in real data rather than self-reported inputs, they trust and act on recommendations.
Amorepacific's generative AI chatbot, called Amore Chat, connects to the company's internal product database and customer information. The tool recommends products, compares options, and summarizes reviews. The bot recently launched on ChatGPT and operates on Amore Mall, the company's online store.
Ongoing personalization replaces one-time recommendations
Static skin routines no longer meet consumer expectations. Brands must monitor skin changes over time and update recommendations as conditions shift with seasons, environment, and lifestyle.
This moves personalization closer to a service model. Brands that encourage regular check-ins and continuous monitoring build returning-user relationships. Those that don't offer these tools risk appearing less capable of addressing individual needs.
Product developers implementing AI systems should focus on two core elements: perceived intelligence and continuity. Systems that combine objective analysis with contextual data and evolve recommendations over time generate higher engagement and trust than static tools.
For teams building these systems, AI for Product Development covers the technical and strategic foundations. Understanding how Generative AI and LLM systems work helps product developers design chatbots and recommendation engines that actually perform.
Your membership also unlocks:
