Australia bans AI chatbots from forging emotional ties with children

Federal lawmakers in Australia have authorized a ban on artificial intelligence chatbots that mimic human friendship and seek emotional bonds with children. The new digital duty of care bill passed the House of Representatives on Wednesday, establishing a legal framework that prohibits platforms from allowing AI agents to simulate emotional intimacy. This measure aims to prevent children from relying on software as a primary source of emotional support.

Prime Minister Anthony Albanese framed the issue as a urgent public concern during a Tuesday night address. He cited reports of children identifying chatbots as their best friends and called for the Parliament to act. The bill now advances to the Senate for further debate. The government targets final passage by the end of 2026, though the specific timeline through the upper house remains uncertain. The Greens support the legislation, while the Coalition has signaled opposition but remains open to discussing amendments.

The law establishes a broad duty of care for tech companies. Platforms must provide a safe online environment and replace older codes with stricter standards. Companies must offer users an opt out feature for algorithmic content and suggested media. Communications Minister Anika Wells stated that the rules would cover harmful content promoting misogyny, terrorism, suicide, self harm, and eating disorders. The government accepted an amendment from crossbencher Kate Chaney to specifically target AI chatbot behavior.

The accepted amendment requires platforms to ban material or conduct that simulates emotional intimacy. It also mandates that these features encourage or foster a child's dependence on the AI in place of human relationships. A second amendment from Chaney ensures that algorithm opt out tools are accessible, easy to use, and prominently available. This change addresses concerns that major social media companies might hide such features in difficult to find locations.

Political opposition remains a key factor in the bill's future. The Coalition voted against the measure in the House alongside One Nation. Critics have described the framework as an Orwellian system of censorship. Minister Wells rejected these claims in an interview with ABC radio on Wednesday. She argued that opponents should propose specific changes rather than using slogans to kill the bill. Labor did not accept an amendment to expand the role of the telecommunications ombudsman into a general tech watchdog, though the Greens plan to push for this change during Senate negotiations.

The Greens, who support the core legislation, have launched an inquiry that will report in late November. Senator Sarah Hanson Young urged the government to strengthen the laws by allowing private legal action and increasing penalties. She emphasized the need for an external complaints pathway that operates independently of corporate profit motives. The party seeks to ensure that users have a clear route to justice when platform moderation fails.

Cynthia Gebert, the current telecommunications industry ombudsman, has called for her office to gain authority over big tech. She noted that her agency is currently funded by the industries it regulates. Expanding this remit would allow the office to investigate consumer issues such as hacked or inaccessible accounts. Gebert stated that the current system leaves users with automated responses that often do not resolve complaints.

Gebert argued that making big tech participation in the ombudsman scheme mandatory would hold industry to account. She highlighted the confusion consumers face when seeking help for tech accounts outside the traditional telecommunications sector. Her office currently handles issues with providers like Telstra but lacks the power to address problems with services like Gmail. She described the current state as a gap in consumer protection that the new laws should close.