The short version
- The House of Representatives approved a digital duty of care bill that bans AI chatbots designed to simulate emotional intimacy with minors.
- Opposition parties rejected the legislation, characterizing it as censorship, while crossbenchers secured amendments regarding algorithm transparency and complaint mechanisms.
- The measure now faces Senate scrutiny, where Greens lawmakers are pushing for expanded legal recourse for users and a dedicated tech watchdog.
Australia’s federal government has advanced significant legislation aimed at regulating online safety, with the digital duty of care bill passing the House of Representatives on Wednesday. The proposed framework introduces a broad obligation for technology platforms to ensure safe environments for users, replacing fragmented existing codes with a unified standard. Central to this new regulatory approach is a specific prohibition against artificial intelligence agents that seek to establish emotional connections with children, reflecting growing governmental concern over the psychological impact of non-human companionship.
Prime Minister Anthony Albanese framed the urgency of the bill around reports of young people treating chatbots as primary confidants. In a speech delivered Tuesday night, he described these interactions as a wake-up call for parliamentarians to act against what he characterized as a replacement of human relationships with simulated ones. The government argues that allowing AI to foster dependence in children poses a distinct harm that current regulations fail to address adequately.
The legislation was amended during lower house debates to include specific language targeting these AI interactions. Crossbencher Kate Chaney’s proposal, which Labor accepted, requires platforms to ban material or conduct that simulates emotional intimacy and encourages a child’s reliance on it instead of human contact. This addition expands the scope of prohibited content beyond traditional categories like terrorism or self-harm, which Communications Minister Anika Wells indicated would also remain under the bill’s purview.
Despite the lower house victory, the path to final enactment remains fraught with political uncertainty. The Coalition opposition voted against the measure, alongside One Nation, accusing the Labor government of establishing an Orwellian censorship apparatus designed to influence online discourse. Minister Wells rejected these characterizations, challenging opponents to propose constructive amendments rather than attempting to defeat the bill entirely. While the Greens support the legislation, their backing is conditional on further strengthening protections.
A key point of contention involves the mechanism for user complaints and industry oversight. Independent MP Sophie Scamps proposed expanding the role of the telecommunications industry ombudsman to include social media and AI platforms, creating an external body where users could seek remedies for issues like account hacking or inaccessible services. Labor declined this amendment in the lower house, and Minister Wells did not endorse the idea publicly. However, the Greens intend to press for this reform during Senate negotiations, arguing that current internal platform processes are insufficient.
Cynthia Gebert, the current telecommunications industry ombudsman, suggested her office is well-positioned to handle such an expanded mandate. She noted that while the duty of care framework places responsibility on platforms to ensure safety, it lacks a robust external avenue for unresolved complaints. Gebert warned that without a dedicated watchdog, users may remain trapped in cycles of automated responses and ineffective escalation procedures. She advocated for mandatory participation by big tech companies in an ombudsman scheme, funded by the industry itself, to drive systemic improvements.
Another accepted amendment from Chaney addresses concerns about algorithmic transparency. The revised bill now mandates that social media platforms provide opt-out features for algorithms and suggested content that are accessible, easy to use, and prominently displayed. This provision aims to prevent companies from burying these controls in obscure settings, ensuring users have genuine control over their digital feeds. The government views this as a critical component of user autonomy within the broader safety framework.
The bill now moves to the Senate, where its fate is less certain. The Greens have initiated an inquiry into the legislation, with findings expected in late November. Senator Sarah Hanson-Young has called for stronger measures, including the right for Australians to pursue private legal action against platforms and increased penalties for non-compliance. With the Coalition opposed and Labor seeking to win over some opposition support through potential amendments, the upper house debates will likely focus on balancing safety mandates with concerns over censorship and regulatory overreach.
The government aims to pass the bill by the end of 2026, but the political dynamics suggest a protracted process. The divergence between Labor’s push for comprehensive duty-of-care standards and the opposition’s resistance highlights deeper ideological divides regarding internet governance. As negotiations continue, the specific definitions of harm, the scope of AI restrictions, and the structure of oversight mechanisms will remain central to the legislative battle.
Sources behind this briefing
Go to the original reporting
- The Guardian World↗‘Emotional intimacy’ from AI chatbots would be banned under Australia’s digital duty of care laws
