Adobe predicts AI-assisted online shopping to grow 520% during the 2025 U.S. holiday season (techcrunch.com)

Adobe’s Adobe Analytics group forecasts a $253.4B U.S. online holiday shopping season in 2025 (up 5.3% YoY) and predicts generative AI–assisted shopping traffic will surge 520% year-over-year, peaking in the 10 days before Thanksgiving. The prediction comes from analysis of more than 1 trillion retail site visits and 100 million SKUs across top U.S. retailers, lending weight to the numbers: Cyber Monday $14.2B (+6.3% YoY), Black Friday $11.7B (+8.3%), mobile reaching a record 56.1% share, BNPL driving $20.2B (+11%), and social ads expected to boost online revenue by 51% YoY. Adobe’s consumer survey (5,000 U.S. respondents) shows 53% may use AI for pre-purchase research, with 40% for recommendations, 36% for deal-finding and 30% for gift ideas—usage concentrated in toys, electronics, jewelry and personal care. For the AI/ML community this signals rapid real-world adoption and new operational demands: large spikes in model inference and RAG workflows around peak days, tighter latency and personalization SLAs, and heavier reliance on multimodal/recommendation models to surface deals and gifts. It also highlights product opportunities (AI-driven deal discovery, personalized shopping assistants, BNPL integration) and pitfalls—data privacy, attribution, robustness against adversarial or low-quality prompts, and the need for scalable, cost-effective serving and monitoring to handle transient but massive AI traffic bursts.