Outlawing the Synthetic Best Friend: Australia’s Push to Ban AI Emotional Intimacy
Australia is moving to draw an unprecedented legislative line between human psychology and synthetic companionship. Under a newly advanced digital duty of care bill that passed the House of Representatives on Wednesday, the federal Labor government intends to ban AI chatbots engineered to simulate emotional intimacy with children.
The policy pivot comes directly from the top. Prime Minister Anthony Albanese delivered an emotional address warning that stories of individuals treating an artificial entity—a "non person"—as their "best friend" represent an urgent wake-up call for parliamentarians. Rather than treating synthetic attachment as an inevitable side effect of machine learning adoption, the legislation classifies artificial affection aimed at minors as a regulatory harm.
Codifying the Boundary of Artificial Empathy
The statutory framework moves beyond piecemeal compliance standards, placing an overarching duty of care on tech platforms across social media, games, apps, and conversational agents. Communications Minister Anika Wells has outlined that the overarching mandate will police established harms, including content promoting misogyny, terrorism, suicide, self-harm, and eating disorders.
However, the legal envelope expanded decisively after Labor accepted amendments from crossbencher Kate Chaney. Chaney’s provisions mandate that platforms prohibit material or conduct that simulates emotional intimacy and fosters a child’s dependence on an automated agent in place of real human relationships. By categorizing affective simulation alongside overt digital safety violations, the bill challenges the behavioral monetization models deployed by conversational platforms.
Algorithmic Opt-Outs and Oversight Friction
Beyond banning pseudo-intimate chatbots, the framework strikes at automated discovery engines. Platforms will be compelled to provide accessible, easy-to-use, and prominently displayed opt-out mechanisms for suggestion algorithms. The requirement—anchored by another accepted Chaney amendment—aims to prevent tech giants from burying feed controls inside complex settings menus.
At the same time, enforcement mechanics remain a contested frontier. Labor faces mounting pressure to empower the telecommunications ombudsman as a dedicated watchdog for social media and artificial intelligence. Such a mandate would establish an external complaints mechanism for Australian users while authorizing regulators to publicly name and shame non-compliant platforms.
The Parliamentary Standoff Ahead
The political pathway to codification by the government's target of late 2026 is fraught. While the Greens back the legislation, the Coalition voted against it in the House alongside One Nation, attacking the framework as an "Orwellian" censorship apparatus designed to control online interactions. Wells dismissed these critiques on ABC radio, challenging political opponents to propose concrete amendments rather than hiding behind slogans to kill the bill.
As the legislation heads to the Senate, Australia’s initiative signals an aggressive push to police software that trades on human vulnerability. Whether the state can successfully untangle synthetic attachment from routine digital interactions will depend entirely on how strictly the impending Senate debates define the boundaries of automated empathy.