Australia Is Taking Choice Back From Tech Platforms, but This Is Only the Starting Point of Algorithmic Accountability

Major tech platforms have long held the power to distribute information and design products, yet they leave the foreseeable risks created by algorithmic recommendations to be borne by children, parents and ordinary users. The "digital duty of care" the Australian government plans to introduce aims to correct this imbalance — not to strip users of choice.

The draft, currently at the consultation stage, requires platforms such as TikTok, Instagram and Facebook to actively ask users whether they accept recommended content, and to allow users to switch to feeds showing only the accounts they follow. For users under 16, platforms must also turn off personalised algorithms and infinite-scroll functions. The reform acknowledges a fact long obscured by the product logic of tech companies: recommendation algorithms are not an unrefusable natural environment imposed on users, but systems actively deployed by platforms, capable of being switched off, and that must bear the consequences.

Platforms hold the technical power to decide how content reaches users, yet ordinary people find it hard to see this arrangement clearly, let alone escape constant recommendation. The risks to children listed in the draft include bullying, pornography, misogynistic content, material that glorifies crime, and content encouraging disordered eating. The duties toward adults focus on material that already constitutes criminal offences: abuse, sexual violence, serious threats, support for terrorist organisations, and encouragement of suicide. Australia's eating-disorder support organisation, the Butterfly Foundation, says it regularly encounters users whose feeds are dominated by idealised, distorted and weight-oriented content, and considers such content one of the important triggers for eating disorders. Bereaved parents and child-protection advocates who have lost children support the legislation, which also shows that platforms' so-called product features have entered the realm of family safety.

The draft proposes fines exceeding AUD 100 million on platforms that fail to meet their obligations — a point of vital importance. Without penalties sufficient to influence corporate decisions, "user autonomy" can easily become a decoration buried in a settings menu. If platforms can design personalised recommendations and infinite scroll, they cannot, once harm emerges, push all the responsibility back onto children and parents who lack the technical capacity to respond.

Critics fear the government will use safety duties to intervene in information content, and that risk must be constrained by legal boundaries. The draft currently limits adult protection to existing criminal offences, and any ministerial decision to add new categories of harm can still be vetoed by the Senate. This article is based on a single Australian Broadcasting Corporation report on the draft and related statements; no platform response and no final legal text are yet available for cross-checking.

What truly needs to be broken is the order in which tech platforms treat algorithmic control as a default prerogative. Australia's plan at least establishes a clear principle: those who choose recommended content should be the users themselves; the companies that design and run the systems must bear the duty to prevent harm; and families should not continue to pay the safety bill for global tech companies.