AI shopping assistants are reshaping holiday retail – and you might already be using one without realizing it (www.techradar.com)

Adobe forecasts a 520% surge in “AI-driven” traffic to U.S. retail sites for the 2025 holiday season—based on analysis of more than 1 trillion site visits—after AI-influenced visits already jumped 1,300% last year. The firm expects $253.4 billion in online holiday spending (a 5.3% year-over-year rise) and finds that over a third of 5,000 U.S. survey respondents have used AI for shopping: 53% use it for research, 40% for product recommendations, 36% for deal-finding, and 30% for gift inspiration. Adobe defines AI traffic as visits that originate from recommendations or links produced by AI interfaces—chatbots, browser sidebars, marketplace autosuggest, or LLM-powered assistants such as ChatGPT and Gemini—so many shoppers may already be guided by AI without realizing it. For the AI/ML community, this cements LLMs and retrieval-augmented systems as frontline recommender engines and discovery layers, shifting importance from SEO/shelf space to LLM relevance and ranking. Technical implications include the need for retailers to expose structured product catalogs, robust APIs, embeddings and metadata for accurate retrieval; new attribution and evaluation metrics to measure LLM-driven conversions; risks around bias, transparency, and monetization (sponsored recommendations vs. neutral ranking); and opportunities to optimize RAG pipelines, prompt design, and multimodal search (image-to-product). As AI referrals drive longer browsing and rising purchases, model explainability, data freshness, and alignment with commercial objectives will directly affect retail performance.