What Happened
New data from fintech platform Adyen found that 100 per cent of surveyed Australian enterprise retailers have used AI in some form, and 95 per cent are familiar with agentic commerce. The figures, reported by CommercialRealEstate.com.au on 4 August 2026, sit alongside record retail property transaction volumes: JLL analysis found that retail property transactions reached A$14 billion last year.
Gary Mortimer, professor of retail marketing and consumer behaviour at Queensland University of Technology, described the change in direct terms. "It really is quite a seismic shift towards a new way that we will actually shop," he said. "What we're currently doing is shopping ourselves, whether that's online or physically in store, but AI will simply enable us to delegate that task."
Jeremy Smith, senior vice-president and head of enterprise ANZ at Salesforce, which developed the AI agent platform Agentforce, drew a clear line between earlier tools and the current generation. "Earlier retail AI and chatbots were basically sophisticated guessing machines," he said. "Agentic AI is an entirely new class of autonomous agents that don't guess, they reason. By utilising a brand's data, sitting within its trusted enterprise environment, an agent acts with deep business context to independently make decisions, solve multi-step problems and execute real work from start to finish."
Why It Matters
The Adyen data points to near-universal AI adoption among large Australian retailers at a time when the technology is moving from recommendation engines toward autonomous purchasing. Where earlier systems suggested products, agentic systems can complete transactions, manage inventory queries, and handle customer service without human input at each step.
Smith said customer service is where deployment is happening fastest. "Customer service is seeing the most explosive immediate impact because it's so easy to deploy agents to handle the routine bulk queries," he said. He also noted broader operational changes: "Marketing is shifting from batch emails to real-time, hyper-personalised journey orchestration. Meanwhile, merchandising and inventory management are being changed by real-time stock visibility that instantly equips store managers and trade partners with data to optimise shelves on the fly."
For property, the question is whether AI-driven shopping reduces the need for physical retail space or changes what that space is used for.
Key Details
Mortimer identified the categories most likely to see early agentic AI adoption. "I think we'll really see it predominantly in food retailing, supermarkets and groceries, and other businesses where the consumption event tends to be quite habitual and routine, as opposed to recreational or social shopping," he said.
He was measured about the implications for physical stores. "I don't think physical retail will change as a function of agentic AI from what it is today, where physical retail's gone through an adaptation due to e-commerce," he said.
The record A$14 billion in retail property transactions recorded by JLL last year reflects a sector that has already absorbed significant structural change from online retail. The additional layer of agentic commerce, which can handle ordering and fulfilment tasks that previously required a consumer to visit a site or a store, adds further complexity to how retailers and property owners plan their physical footprints.
Background and Context
Agentic AI is a branch of artificial intelligence in which an agent acts independently with minimal human intervention, distinct from generative AI tools such as ChatGPT or Claude that respond to prompts but do not act autonomously. Its decision-making capabilities make it suited to shopping tasks: ordering groceries, comparing products, and completing purchases on behalf of a consumer.
Salesforce's Agentforce is one commercial platform in this space. Smith's description of the technology, grounded in enterprise data environments, reflects the current industry framing: agents are positioned as operating within a brand's own systems rather than as general-purpose consumer tools.
Australian retailers' near-universal AI experimentation, as captured in the Adyen survey, suggests the infrastructure for broader agentic deployment is already in place at the enterprise level.
What Comes Next
Mortimer's view is that the clearest near-term effects will appear in grocery and other routine consumption categories. Recreational and social shopping, which draws consumers to physical centres for reasons beyond the transaction itself, is seen as more resistant to agentic substitution.
JLL's record retail property transaction figure points to continued investor interest in the sector despite, or alongside, the technology shift. How logistics property demand evolves as agentic systems take on more fulfilment-related tasks remains an open question that property analysts and retailers are watching.