Australia’s digital duty of care bill has moved one step closer to becoming law, with the House of Representatives passing legislation that would bar AI chatbots from building emotional intimacy with children. The bill passed on Wednesday, 28 October, and now goes to the Senate, where its final shape will depend on amendments, committee findings and political negotiations.

Key facts

• The House passed the digital duty of care bill on Wednesday, 28 October; it now goes to the Senate.
• Amendments from independent MP Kate Chaney, accepted by the government, add certain chatbot conduct to the bill’s list of harms.
• Platforms would have to ban material or conduct that simulates emotional intimacy and encourages or fosters a child’s dependence on it in place of human relationships.
• The wording covers children’s dependence on a chatbot; it does not cover adults who use AI companions.
• Prime Minister Anthony Albanese urged the opposition Coalition to support the bill, saying stories about people treating a chatbot as a best friend should be a wake-up call.
• The framework sets online safety standards for social media, games, apps and chatbots and replaces several existing codes with a broader duty of care.
• The bill also requires social media platforms to offer an algorithm opt-out that is accessible, easy to use and prominently available.
• The Coalition and One Nation voted against the bill in the House; opponents called it Orwellian censorship, a claim the communications minister rejected.
• The Greens support the bill, have set up a Senate inquiry due to report in late November, and want a right to sue platforms plus higher penalties.
• Independent MP Sophie Scamps proposed expanding the telecommunications ombudsman into a complaints body for social media and AI; Labor did not accept it in the lower house, and the Greens will push it in the Senate.

A narrow rule aimed at a broad concern

The chatbot provision is carefully drawn. It does not seek to ban AI companions for adults, nor does it attempt to regulate every conversational agent. Instead, it targets a specific design and use case: systems that simulate emotional intimacy and encourage or foster a child’s dependence on the chatbot in place of human relationships. That language puts the focus on the relationship between the child and the machine, not simply on the content a child might encounter. It also gives regulators a hook to examine whether a product is designed to keep a child engaged in a way that substitutes for family, friendship or other human connection.

The amendment was introduced by independent MP Kate Chaney and accepted by the government on the same day the bill passed the lower house. Its inclusion reflects growing concern about anthropomorphic AI products that speak in friendly voices, remember personal details and present themselves as always available. For children, those features can be especially compelling. A chatbot can offer reassurance at any hour, never appear bored and never push back in the way a friend or parent might. Those qualities may be useful in some contexts, but they can also blur the line between a tool and a relationship.

Albanese’s appeal and the government’s case

Prime Minister Anthony Albanese made the case for the bill in a speech on Tuesday night, again asking the opposition Coalition to vote for it. He pointed to stories about people identifying a chatbot, a non-person, as their best friend and said such accounts should be a huge wake-up call for every member of parliament. The remark framed the legislation as a response to a cultural and technological shift, not merely a technical update to online safety rules.

The broader framework would set new online safety standards for social media, games, apps and chatbots. The government says it places an overarching obligation on platforms to keep Australians safe online. It would replace several existing online codes and standards with one broader duty of care. That structure is meant to be more flexible than a patchwork of narrow rules, allowing regulators to respond as products change. It also shifts more responsibility onto platforms to assess and mitigate risks, rather than waiting for harmful content to spread before acting.

The bill also makes social media platforms offer an algorithm opt-out. Another Chaney amendment, also accepted, says that opt-out must be accessible, easy to use and prominently available. That requirement matters because a right that is buried in settings is unlikely to be used. Communications Minister Anika Wells has said the laws would capture content that promotes misogyny, terrorism, suicide, self-harm and eating disorders. Those categories sit alongside the chatbot provision, showing that the bill is intended to cover both harmful content and harmful design.

Opposition and the censorship argument

The Coalition and One Nation voted against the bill in the House. Opponents have accused Labor of building an Orwellian system of censorship, claims Wells has rejected. She said that if critics really care, the test is simple: they should bring an amendment, not a slogan to kill the bill. The exchange captures the central political dispute. Supporters see a duty of care as a baseline safety measure, similar to obligations in other consumer products. Critics see a broad regulatory power that could chill speech or impose vague standards on platforms.

The chatbot amendment may intensify that debate. Its language is not a simple content ban. It asks platforms to identify conduct that simulates emotional intimacy and fosters a child’s dependence. That requires judgments about design, context and age. Supporters say such judgments are already made by platforms when they optimize for engagement. Critics may argue that the wording is open to interpretation and could capture products that are not obviously harmful. The Senate process is likely to test how narrowly the provision can be written while still addressing the concern.

The Senate fight

The Greens support the bill and have set up a Senate inquiry, due to report in late November. They want Australians to be able to sue platforms and want higher penalties. Labor hopes the Senate passes the bill by the end of 2026 and says it is open to amendments. That timetable leaves room for changes but also creates pressure to keep the coalition of support together. The inquiry will give health professionals, child safety advocates, industry groups and civil liberties organizations a chance to argue over definitions, enforcement and penalties.

The Senate is also where the ombudsman proposal is likely to resurface. Labor is under pressure to make the telecommunications industry ombudsman a complaints body for social media and AI. Independent MP Sophie Scamps proposed the change, which Labor did not accept in the lower house. The Greens will push for it in the Senate. The ombudsman, Cynthia Gebert, said her office could take on the role, paid for by big tech. She illustrated the gap by noting that her office can help with a phone account but not an email account. The point is that Australians have a clear place to complain about some communications services, but not about the social media and AI services that increasingly shape daily life.

Why the chatbot rule matters

The chatbot rule follows concerns abroad about AI companions and children. Australia already has an under-16 social media ban, and the duty of care bill is part of a wider push to regulate online spaces more aggressively. The focus on emotional intimacy is significant because it moves beyond the usual categories of bullying, explicit content or screen time. It asks whether a product is designed to become a substitute for human relationships, especially for children who may be more vulnerable to anthropomorphic design.

There is also a practical question of enforcement. Platforms would need to know whether a user is a child, which raises age assurance issues. They would need to assess whether a chatbot’s tone, memory features, prompts and rewards amount to simulated emotional intimacy. They would need to show that they have taken steps to prevent dependence in place of human relationships. Those are not easy tests, and the bill does not spell out every technical detail. Much will depend on how the regulator interprets the duty and what evidence it demands.

The bill’s supporters argue that uncertainty is not a reason to wait. AI companions are already available to children, often through apps and games that do not look like traditional social media. The risks may be subtle: a child withdrawing from friends, preferring a chatbot’s constant approval, or treating a machine as the only reliable listener. The amendment’s wording tries to capture that harm without banning all chatbot use. Whether it succeeds will depend on the Senate, the regulator and the courts.

The Senate inquiry is due to report in late November. Labor hopes to pass the bill by the end of 2026. The final text may change on penalties, rights to sue and the ombudsman role. The chatbot provision, however, has already become one of the bill’s most closely watched features, because it addresses a question that lawmakers elsewhere are only beginning to confront: when does a machine become more than a tool in a child’s life?

Source:TNW | Au News