AI promises a faster search for materials that balance performance, safety, cost and sustainability, but incomplete data can weaken the answer. Where can predictive tools genuinely shorten development? And what evidence is needed before brands trust their recommendations on formulation, environmental impact, and regulatory screening? • How are AI and predictive modelling changing the search for alternatives to fossil-based ingredients and packaging? • Can AI reduce the physical experiments needed to balance barrier performance, durability, safety, and sustainability? • How reliable are AI-enabled life-cycle assessment and regulatory screening, and where is lab validation still needed?