Australia's duty of care bill now covers AI chatbots, so a bot that gets too friendly with kids becomes a legal problem
By Brett Gavaghan, Founder of Adaptd ·
On 28 October the House of Representatives passed the Online Safety Amendment (Digital Duty of Care) Bill, which brings AI chatbots under one broad online safety duty. It now goes to the Senate.
The short version
- Australia's Digital Duty of Care bill puts one broad online safety duty on platforms, and now names AI chatbots.
- It passed the House of Representatives on 28 October 2026 and now goes to the Senate, with an inquiry due in late November.
- If your business runs its own chatbot or app that young people might use, ask whoever built it how it would meet this duty.
Most of Australia's online safety rules were written with social media in mind. This bill writes AI chatbots into the same rulebook, with a specific rule about children.
What is the Digital Duty of Care bill?
The Online Safety Amendment (Digital Duty of Care) Bill 2026 replaces several existing online safety codes with one overarching duty to keep Australians safe online. It passed the House of Representatives on 28 October 2026.
The duty covers social media, online games and apps, and now AI chatbots as well. An amendment from independent MP Kate Chaney, accepted by the government, adds a specific chatbot harm: material or conduct that simulates emotional intimacy and encourages or fosters a child's dependence on it. That rule protects children, not adults using AI companions.
When would it start, and what are the penalties?
The bill isn't law yet. It still has to pass the Senate, where an inquiry is due to report in late November. The government hopes to pass it by the end of 2026.
Reports put the maximum penalty at A$109.2 million. The duty would start after a 12-month transition following royal assent, so late 2027 at the earliest. Plenty of time, which is usually how these things sneak up on you.
What should your business do about it?
Most businesses that simply use an AI assistant like ChatGPT at work won't need to do anything, because the duty falls on the services themselves. The ones to watch are businesses that run their own chatbot or app.
My take: if you've built a customer chatbot, ask your developer or vendor two questions now. Could a child end up using it, and how does it handle someone who keeps coming back to it for comfort? If the answers are vague, get proper legal advice before the Senate finishes.
Too small to worry? The duty is written for online services, not only big tech, so check rather than assume. A polite, slightly boring bot is starting to look like the safest design brief going.
Sources
Wondering what this means for your business?
That's exactly what the discovery call is for.