On September 11, 2026 the Online Safety Amendment (Strengthening Enforcement for the Social Media Minimum Age) Bill 2026 received assent in Australia, expanding the eSafety Commissioner's information-gathering powers and lifting maximum civil penalties for non-compliance with the social media minimum age framework to A$99 million (about US$71 million), according to Biometric Update. The bill was introduced in June and amends the Online Safety Act 2021. eSafety Commissioner Julie Inman Grant had publicly pushed for the authority to compel platforms to produce documentation rather than rely on platform self-reporting; she now has it.

Why It Matters

Australia has been the global reference implementation for age assurance, and this is the next phase: from "verify age at the door" to "be structurally responsible for what happens inside." For the sexual technology and adult content sectors, three elements of the draft matter directly. First, the duty explicitly attaches to app distribution services and hosting providers — the chokepoints that already determine whether sex tech apps and adult platforms can reach Australian users at all. Second, the nudification takedown power is one of the first statutory hooks aimed squarely at the AI undressing app category rather than at deepfakes generically. Third, the risk-assessment-and-mitigation structure imports the EU Digital Services Act's compliance model into a jurisdiction that has shown far more appetite for enforcement. The criminal exposure for individuals is the genuinely novel bit. Civil penalties are a cost of doing business for companies with Meta's balance sheet; a personal liability attaching to a local managing director is not. Expect the compliance function at Australian subsidiaries to get considerably more assertive, and expect the same drafting to be studied by legislators in the UK, Ireland and Canada. For smaller sex tech and adult operators, the practical risk is the familiar one: rules written for platforms with thousand-person trust and safety teams land hardest on companies with none. It is worth noting the duty of care instrument is an exposure draft, not law — the consultation outcome will determine how much of this survives.

Three days earlier, on September 8, the federal government released the exposure draft of a second and much broader instrument: the Online Safety Amendment (Digital Duty of Care) Bill 2026. As summarised by law firm Lander & Rogers, the draft would create a statutory duty requiring online service providers to ensure a "safe online environment" as far as reasonably practicable — an obligation that reaches well beyond social media to search engines, app distribution services, hosting providers, and AI content and generation services. Providers would have to conduct written risk assessments identifying reasonably foreseeable harms, implement measures to address them, and supply "user empowerment tools" letting people control how design features operate.

The draft names the design features it distrusts: recommender systems, infinite scrolling, engagement metrics such as likes and views, and disappearing content. For Australian users under 16, social media services would have to ensure those features do not operate at all. The exposure draft lists pornography first among the categories of content it treats as harmful to children. It also proposes powers to require app stores and search engines to remove apps or websites predominantly used to generate fake nude material, to issue link deletion notices to search engines, and to order removals within 24 hours.

The penalty architecture is where it gets sharp. Lander & Rogers reports maximum civil penalties of 60,000 penalty units per breach of the duty — at the current A$364 penalty unit value, over A$21.84 million per contravention — and 6,000 units for transparency, complaint-handling and link deletion failures. Separately, Mumbrella reported on September 9 that individuals who refuse to provide information, produce documents or take an oath during an eSafety examination could face up to 12 months' imprisonment or a civil penalty of A$364,000, double the originally proposed figure. Mumbrella attributes the doubling to the Greens, whose support the bill needs in the Senate, and notes the provision applies to any individual holding relevant information — which in practice includes senior local executives at Meta, Google and TikTok.

Communications Minister Anika Wells has called the package a "reckoning" for US tech companies. The context is a minimum-age regime whose early results both sides are spinning: more than 80% of Australian under-16s were still using social media three months after the law took effect, but under-16 account ownership fell over 10% in the same window, and more than half of children who kept their accounts said they were simply never asked to prove their age.

Sources

Update — 2026-09-14

{Initial entry — story first created.}

Update — 2026-10-14

The bill heads to parliament this week, with one change already conceded and a sexual-health problem in its definitions. ABC News reported on October 12, 2026 that the Albanese government will introduce the Digital Duty of Care legislation this week and is "prepared to scrap" the ministerial discretion that would have let the minister add new categories of harm by regulation. The Coalition had attacked that power as a route to political censorship. The duty would require social media, search, AI chatbots and other platforms to take reasonable steps against foreseeable "harms" to minors and "serious harm" to adults, and would give users the option to switch off algorithmic feeds. The government says the bill reflects 225 submissions, 22 roundtables and 60 one-on-one meetings.

The Senate math runs through the Greens. Senator Sarah Hanson-Young has set three tests: a private right of action so people harmed online can sue platforms for compensation, fines tied to a percentage of global revenue rather than the roughly A$100M cap, and measures that address addictive design. The Coalition still opposes the bill.

The sex-ed problem. In The Conversation (published October 14 AEDT), public-health researchers point out that one section defines material that "encourages, promotes, urges or instructs" illicit drug use and some kinds of sexual activity as "seriously harmful" to adults as well as young people, with no carve-out for harm reduction. They note that platforms already remove this kind of content. NAPWHA's "Can't Pass It On" HIV campaign was taken down from Facebook and YouTube as "misinformation," and Meta bans sexual and reproductive health product ads unless they are deemed to target adults. A statutory duty with no harm-reduction exemption gives platforms a legal reason to over-moderate. That affects sex educators, STI campaigns, PrEP providers and sexual-wellness brands, which already depend on coded language to get through moderation.

Why it matters now: the final text matters more than the headline. If the "sexual activity" definition passes unchanged, Australia will have written into law the over-moderation the industry already deals with informally, and the EU-style risk-assessment model means platforms will apply it conservatively. Watch for a harm-reduction amendment in committee, and for whether the Greens' private-right-of-action demand survives the trade-off for their Senate votes.

New Sources