Australia's Competition and Consumer Amendment (Unfair Trading Practices) Act 2026 received Royal Assent on July 6, 2026, inserting a general prohibition on unfair trading conduct into the Australian Consumer Law. The statutory test is deliberately abstract: conduct that manipulates a consumer, or unreasonably distorts the environment in which a consumer makes a decision, and causes or is likely to cause detriment. Nowhere does the Act mention artificial intelligence, algorithms, or automated personalization. That silence is the point — and it is also the article's central tension.
The case for going technology-neutral
The strongest argument for this approach is straightforward: naming a technology invites obsolescence. A statute that lists "algorithmic pricing" or "AI chatbots" as prohibited tools would already be outdated by the next model release, while conduct-based drafting — did the trader manipulate the consumer's decision, regardless of the mechanism — survives technological change. The Australian Competition and Consumer Commission (ACCC) named exactly this problem in its February 19, 2026 announcement of 2026-27 enforcement priorities, with Chair Gina Cass-Gottlieb flagging "subscription traps and other dark patterns that manipulate consumer behaviour" as a top-line concern for the coming year, well before the Act cleared Parliament. The regulator is not waiting for AI-specific legislation to act on AI-adjacent harms — it is repurposing consumer law's oldest tool, the ban on unfair conduct, for the newest delivery mechanism. That is a defensible bet: subscription cancellation flows, drip-priced checkout screens, and confirm-shaming pop-ups have manipulated consumers for two decades without any AI involved at all, and a growing body of ACCC casework already treats them as harmful regardless of how they're built.
Where technology-neutral drafting creates real cost
The trade-off is legal certainty, and it falls hardest on exactly the AI deployments Australia says it wants to encourage. "Unreasonably manipulates" and "unreasonably distorts the decision-making environment" are judgment calls, not bright lines. A firm running large-scale AI-personalized offers — dynamic pricing, tailored upsell timing, adaptive UI nudges — cannot easily pre-clear that conduct against a standard that will be interpreted case by case, starting only when the ACCC brings its first enforcement action after the law takes effect on July 1, 2027. Compare this to the EU's Digital Services Act and Unfair Commercial Practices Directive, which at least enumerate dark-pattern categories (interface interference, forced continuity) that businesses can audit against, or the UK's Digital Markets, Competition and Consumers Act 2024, which paired its own unfair-practices reform with detailed CMA guidance well ahead of commencement. Australia's approach defers that specificity to enforcement discretion. For a small SaaS company running standard AI-driven onboarding flows, the honest answer to "is this compliant?" may not exist until a regulator says so after the fact.
The penalty exposure is genuinely large
This uncertainty would matter less if the stakes were modest. They are not. Corporate contraventions of the Australian Consumer Law now carry penalties of the greater of $100 million per contravention, three times the benefit obtained, or 30% of adjusted turnover during the breach period — a doubling of the prior $50 million cap that the government has openly framed as making penalties "a serious offence with serious consequence" rather than a line item, per Assistant Minister Andrew Leigh's own framing of the broader penalty agenda. Individuals face a $2.5 million ceiling. For a large platform with AI-personalized checkout flows across millions of Australian users, a single adverse finding under an untested "manipulates" standard could be enormously expensive — precisely when the standard itself has never been litigated.
What proportionate enforcement should look like
None of this argues against the underlying policy goal. Subscription traps and drip pricing genuinely harm consumers, and a principles-based ban that doesn't need constant amending as manipulation techniques evolve is sound statutory design — legislators cannot enumerate every dark pattern that will exist by 2030. But proportionate implementation of a broad standard requires the regulator to narrow it through guidance, not litigation. The twelve months between Royal Assent and the July 2027 commencement date is exactly the window in which the ACCC should publish safe-harbor examples distinguishing lawful AI personalization from prohibited manipulation, the way it has already flagged dark patterns and subscription traps as compliance priorities. Absent that clarity, the law's technology-neutral design — its chief structural strength — risks functioning as a technology-neutral chilling effect, deterring legitimate AI-driven commerce alongside the manipulative conduct it was built to catch. Australia should use its lead time well; a law this broadly written earns its legitimacy only through how narrowly it is ultimately enforced.