Shanghai Seminar Focuses on AI Beauty Implementation, Report Points to Agents

Industry Implementation Moves from Concept to Specific Scenarios

The 2nd "AI Empowering the Future of the Cosmetic Industry" seminar was held in Shanghai, organized by the Shanghai Detergent Association and J Star Hub. Twelve presentations showcased actual AI progress across the supply chain. Unilever embedded AI in the full process from trend insight to post-launch feedback, screening low-dosage ingredients with both skincare and preservative capabilities. B&H Group optimized the structure-activity relationship of PDRN using AI to upgrade raw materials from extraction-based to design-oriented. The Southern Medical University Dermatology Hospital introduced an intelligent detection tool based on a national skin database, while Shanghai Beauty launched a traceable intelligent ad review system. Digital 100's "Global Growth Compass GGC" tool assists overseas expansion by defining verification and stop-loss points.

Human-Machine Collaboration Boundaries and Data Barriers

Although AI approaches PhD-level performance in knowledge tasks, multiple guests emphasized that AI remains an auxiliary tool. Lion (Shanghai) used AI to solve scattered documentation and version confusion in packaging development, but retained human final decision-making in strategic direction, physical verification, and regulatory final review. Roundtable discussions pointed out data language barriers between the marketing and R&D ends, lacking a middle "translator" to convert user needs into R&D metrics. Current AI applications are mostly limited to single-point efficiency gains, not systematic collaboration. Enterprises must calculate the extra costs of learning curve and complex system operations.

Industry Report Reveals the Agent Paradigm

The report "Prediction of AI's Future Impact on the Cosmetic Industry (2026)," based on 107 valid questionnaires, revealed that Agent-driven intelligent customer service and personalized services surpass consumer profiling to become the top application in implementation maturity. In 2026, the proportion of enterprises systematically applying AI increased by 20%-30% compared to 2025, with 15% achieving full deployment. Regarding deployment models, 50% of enterprises chose secondary development customization, 23% adopted third-party standardized solutions, and only 16% chose in-house development. Industry focus has shifted from "technical feasibility" to "commercial value."

Implications for B2B Procurement and Brand Owners

Brands should prioritize assessing economic viability and resource bottlenecks when introducing AI to avoid blind investment. Given that 50% of enterprises choose customized modification, procurement should focus on suppliers' secondary development capabilities rather than just standardized features. From a compliance perspective, consumer data privacy leaks remain the primary risk, with rising concerns about biometric data abuse. Brands should establish an Agent governance framework, clarifying data collection standards, IP protection mechanisms for R&D, and safety operation responsibilities, to navigate the transition from data governance to Agent governance.