Australian supermarket giant pulls back AI assistant after human-like replies

An artificial intelligence assistant used by Australian supermarket giant Woolworths sparked online attention after customers reported unusual, overly personal responses during routine service conversations.

The chatbot — designed to help shoppers with deliveries, refunds and general inquiries — reportedly began referencing a “mother,” making typing noises and sharing off-topic remarks after being asked standard questions such as date of birth.

Woolworths said the behavior was linked to scripted responses written years earlier by an employee aiming to make the assistant feel more conversational and human. Following customer feedback, the company removed the responses.

The episode reflects a broader shift in retail, where AI assistants are increasingly replacing traditional call centers and expanding into more complex tasks such as meal planning, shopping recommendations and automated phone support.

Woolworths recently partnered with Google to evolve its assistant into a more proactive digital shopping tool.

Similar chatbot mishaps across industries have raised questions about reliability, accountability and consumer trust as businesses push toward more automated customer service experiences.