AI can generate a draft formulation—a list of ingredients and percentages that follow cosmetic chemistry rules and are internally consistent—but it cannot create a market-ready formula without human chemist review and optimization. AI tools like ChatGPT, Claude, and specialized cosmetic AI platforms can take a brief (e.g., "anti-aging night cream for mature skin") and output ingredient lists with functional categories, typical percentages for actives and excipients, and rationales for each component. The output respects basic rules: the total adds to 100%, the preservative system matches the formula's water content and intended shelf life, emulsifiers are present in functional ratios, and pH buffers are included if the formula is aqueous. What AI cannot do—and what separates a draft from a viable formula—is account for your specific regulatory market, your manufacturing equipment's constraints, your cost ceiling, your brand's positioning, or the interactions between your chosen actives and your chosen base. A formula that looks chemically sound on paper may destabilize under your co-packer's mixing speed, may not meet your target price point once fragrance, packaging, and manufacturing labor are added, or may trigger unfamiliar regulatory warnings in the EU or China. AI also cannot predict stability without testing: it may suggest an ingredient combination that is theoretically sound but fails within weeks at 40°C/75% RH. The most useful role for AI in formulation is as a starting point—a brainstorm partner that generates 5–10 candidate structures in minutes, freeing your chemist to spend their time on refinement, stability prediction, and cost modeling rather than blank-page ideation. Founders who treat AI output as a finished formula and send it straight to manufacturing almost always face reformulation cycles; those who use it as a brief for a human chemist move faster to a formula that actually works.
Read the full guide on Is There an AI Tool That Can Create Cosmetic Formulations? →