Two deadlines will converge on European retail in summer 2026.
Effective July 1, the EU will apply a fixed €3 customs duty on consignments under €150, a change covering 93% of e-commerce flows into the EU.
Beginning in August, EU AI Act transparency obligations will come into force, where online marketplaces must inform users when they are interacting with an AI system.
Both deadlines are set to transform delivery terms, according to a study released by nShift.
“We’re watching a new selection layer emerge,” says Jurgen Leijdekker, CEO at nShift. “Customers may choose brands, but their shopping journey today is increasingly mediated by AI. To achieve e-commerce success, retailers’ delivery offers must be easily comparable. Retailers whose delivery promises that are clear to both human and AI shoppers, and which they live up to consistently, will be well-placed to win.”
Key takeaways:
· AI is automating e-commerce. AI agents are increasingly comparing prices, delivery dates and return terms before shoppers visit the site.
· Post-purchase pages are becoming commercial inventory. Customers spend more time on order-tracking pages and messages. Sales offers made here are more likely to convert and retailers could market these opportunities to brands. IAB Europe forecasts that European retail media spend could reach €31 billion by 2028.
· Out-of-home delivery is now mainstream in Europe. 35% of shoppers use lockers or parcel shops for delivery and 79% prefer to return via locker or parcel shop, according to DHL.
- Customs reform and trade volatility are landing in the customer experience. With EU customs changes and EU AI Act transparency requirements, retailers will need landed-cost clarity, tracking consistency, and auditable delivery data
“The practical challenge is making this work across hundreds of carriers and dozens of markets,” says Mattias Gredenhag, CPO at nShift. “Retailers need delivery and return terms that are machine-readable, carrier events that map into one tracking story, and governance that stands up to scrutiny. The technology is ready. The hard part is making the underlying data consistent.”
