The internet giant revealed plans to add AI Mode to its search engine in another 40 countries including Australia on Wednesday after testing the technology in the US.
Technology experts say the addition could prove useful in low-stakes searches, although they warn users should remain sceptical about its results.
The addition will arrive five months after its announcement at its developers' conference in May, and after AI rival OpenAI revealed plans to establish an Australian base to secure more local partnerships.
When it launches this week, Google's AI Mode will appear as a menu option in its search engine and will use a customised version of the company's Gemini 2.5 generative AI model to produce results.
Users will be encouraged to ask extended and complicated questions using natural language, and Google product management vice-president Hema Budaraju said American users had embraced the challenge.
"Early testers of AI Mode are asking much longer queries - almost two to three times the length of traditional searches," she told AAP.
"It reduces the barrier to getting to that kernel of information that you're seeking and opens up a greater breadth of information from across the web."
As with generative AI tools, users could ask Google's AI Mode to create customised travel itineraries or compare coffee brewing methods for the search engine to deliver a report with web links revealing its sources.
The software addition would be available as well as the AI Overviews shown on top of some existing web searches, Ms Budaraju said, which had proven popular with many users.
"We know how popular AI overviews are because they're now driving over 10 per cent more queries globally ... and we are finding that this growth continues to increase over time," she said.
"You should be able to ask anything and get help from search."
Google's AI web search summaries had also regularly shown incorrect information to users, University of the Sunshine Coast computer science senior lecturer Erica Mealy said.
While some like obviously incorrect advice about adding glue to pizza to make the cheese stick were funny, she said, other AI mistakes could be harmful, inconvenient or embarrassing.
Identifying potential "edge cases" before they happened would be hard even for large tech firms, Dr Mealy said, so anyone who used AI tools should scrutinise results carefully and use them for "low-stakes" advice rather than health or financial advice.
"As long as we're clear that it's an AI overview and we keep our sense of scepticism about it ... it could be helpful," she said.
"It is on us as the users to make sure it's helpful, not harmful."
17°C