This article was co-authored with Lee-Oz Lifschitz.

The rapid rise of artificial intelligence (AI) brings great opportunity but is also testing the limits of competition and consumer law regimes around the world. With the technology still in its relative infancy and the market still contestable, policymakers are emphasising the importance of anticipating and proactively addressing regulatory challenges as they emerge. This article explores three live issues for Australian competition and consumer law in the AI space, which highlight the Government’s appetite for targeted reforms to address specific risks, and a broader strategy of ongoing monitoring to stay ahead of risks that have yet to crystallise.

Targeted regulation of ‘dark patterns’ through prohibition against unfair trading practices

The design of online consumer experiences to influence or manipulate behaviour, known as ‘dark patterns’, has been said to represent a potential gap in the Australian Consumer Law (ACL) framework. See, for example, the ACCC Digital Platforms Services Inquiry Final Report (March 2025, released 23 June 2025), pp 179-84. Accessible at: https://www.accc.gov.au/about-us/publications/serial-publications/digital-platform-services-inquiry-2020-25-reports/digital-platform-services-inquiry-final-report-march-2025. The concern was that these practices harm consumers by interfering with or limiting choice, through tactics such as selectively presenting information or creating a false sense of urgency at checkout, OECD Publishing, Dark Commercial Patterns, OECD Digital Economy Papers No. 336, October 2022. Accessible at: https://www.oecd.org/en/topics/sub-issues/dark-commercial-patterns.html. but may not clearly meet the statutory requirements for misleading or deceptive conduct or unconscionable conduct. Various calls have been made to regulate dark patterns through targeted legislation, for example as part of a long-sought general prohibition against unfair trading practices (UTPs) in the ACL.

On 2 July 2026, the Competition and Consumer Amendment (Unfair Trading Practices) Bill 2026 (UTP Bill) passed both houses of Parliament, and will commence on 1 July 2027 On 25 June 2026, the Senate Economics Legislation Committee recommended that the UTP Bill be passed. The Committee’s report is available here. https://www.aph.gov.au/Parliamentary_Business/Committees/Senate/Economics/UnfairTradingBill2026 (see our update here). The UTP Bill added to the ACL a prohibition against business conduct that causes (or is likely to cause) detriment by either: The UTP Bill also contains specific provisions addressing ‘drip pricing’ and subscription practices.

  • manipulating a consumer; ~~The proposal in the Exposure Draft of the bill for the prohibition to target conduct that unreasonably manipulates a consumer was not included in the Bill.~~or
  • unreasonably distorting the environment in which a consumer makes decisions.

The explanatory materials accompanying the UTP Bill make clear that this prohibition is intended to capture practices that ‘nudge or pressure consumers into unintended actions’ such as through the use of unnecessary countdown timers or overly complex interfaces to overwhelm users – that is, dark patterns. Commonwealth of Australia, Explanatory Memorandum for the Competition and Consumer Amendment (Unfair Trading Practices) Bill 2026), [1.30]-[1.33].

While a potential UTP law had been under consideration since before the AI surge, the rapid and widespread uptake of AI technology in recent years would no doubt have been a significant motivator for the Government in pursuing the reforms. Indeed, an industry snapshot into the AI sector conducted by the Australian Competition and Consumer Commission (ACCC) in December 2025 highlighted the potential for AI to ‘supercharge’ the potency of dark patterns, with technologies capable of producing dynamic content that reacts to users in real time. ACCC, Recent developments in artificial intelligence: industry snapshot, December 2025. Accessible at https://www.accc.gov.au/about-us/publications/recent-developments-in-ai-industry-snapshot.

The UTP Bill represents a major change in ACL compliance requirements for all businesses, but especially those operating online and looking to integrate AI into their platforms. Businesses will need to consider the consumer experience in a much more holistic way. This will mean looking at whether information provided is complete and accurate, and the general design of their user interface and the risks of unduly interfering with user choice. We can expect the ACCC to continue to closely monitor the use of dark patterns across the digital economy, particularly where AI tools are employed, as highlighted by the ACCC’s emphasis on manipulative practices as a key enforcement priority for 2026-27. See ACCC 2026-27 compliance and enforcement priorities, announced 19 February 2026. Available at: https://www.accc.gov.au/about-us/accc-strategy-and-priorities/compliance-and-enforcement-priorities. For more on the significant increase in penalties for ACL breaches recently introduced, see our article here.

Beyond dark patterns, AI raises other novel issues for competition and consumer law, discussed in the next two sections.

Spotlight on agentic AI

The ACCC has noted the rapid pace of development of agentic AI and its potential economy-wide impact. ACCC Snapshot, pp 18-19.

‘Agentic AI’ generally refers to a type of AI that performs tasks autonomously, including running its own internet searches and even controlling other computer programs. The rapid evolution of this technology is emblematic both of the dynamic nature of the market and the challenges lawmakers and regulators face in keeping pace. With each quantum leap in the technology’s capabilities, novel competition and consumer law issues are likely to arise.

One such issue is the attribution of liability for AI agents that breach competition or consumer law. Many existing mechanisms for determining a corporation’s liability under the Competition and Consumer Act 2010 (Cth) (CCA) involve imputing human intention or conduct onto that corporation. For example, section 139B of the CCA provides that in cases where it is necessary to prove a corporation’s state of mind, it can be established through a director, employee or agent of the corporation acting within scope of their authority. The uptake of agentic AI could significantly reduce human involvement in a company’s day-to-day activities, creating a significant grey area for corporate liability under the legislation.  While this issue has yet to be tested in an Australian court, the high-profile Canadian case of Moffatt v. Air Canada is instructive. In Moffatt, the Civil Resolution Tribunal held that Air Canada was liable for misleading statements made by its AI chatbot in the same way as for the contents of static pages on its website. The case garnered significant international attention as an early example of a judicial body grappling with increasingly autonomous AI tools. The Tribunal rejected Air Canada’s submission that the chatbot’s level of autonomy marked it as a separate legal entity responsible for its own actions, and it is conceivable that an Australian court would have reached a similar conclusion on those facts.

But future developments in agentic AI could test the limits of existing liability attribution mechanisms. Indeed, Treasury cautioned in its recent review of AI and the ACL that the advent of agentic AI may soon necessitate revisiting those issues to ensure the ACL framework generally remains fit for purpose. Commonwealth Treasury, Review of AI and the Australian Consumer Law (Final Report, October 2025) 25. The ACCC has similarly signalled the need for ongoing monitoring of potential competition law risks from agentic AI, such as AI agents colluding with each other or engaging in steering or self-preferencing practices.ACCC AI industry snapshot, 21-22. A white paper on agentic AI released by the Governance Institute of Australia in May also highlighted the need for vigilance at an industry level, emphasising the importance of robust governance and risk management strategies to respond to competition and consumer law challenges as they emerge. Governance Institute of Australia, White Paper on Governance in the age of agentic AI, May 2026. Accessible at https://www.governanceinstitute.com.au/advocacy/governance-in-the-age-of-agentic-ai/. It may be some time yet before the regulatory landscape fully takes shape, but it is clear that policymakers and industry players, confronted with the speed of agentic AI’s advancement, are looking to map out that landscape as early as possible.

Strategic partnerships, acqui-hires and competition enforcement

The Government released its National AI Plan in 2025, setting out ambitious goals for growing Australia’s AI industry, including attracting investment in critical infrastructure such as data centres. Department of Industry, Science and Resources (Cth), National AI Plan (December 2025). Accessible at https://www.industry.gov.au/publications/national-ai-plan. (For comprehensive guides to the legal and regulatory landscape governing data centres and digital infrastructure in Australia, see our page here.)  The ACCC has flagged that one of its key priorities will be to guard against entrenchment and interdependency across the AI supply chain.  As such, we can expect the ACCC to closely scrutinise AI-related mergers – which, given the size of the players in the industry, are almost certain to trigger the thresholds under Australia’s new mandatory and suspensory regime and will be likely candidates for Phase 2 review (for a detailed explainer of the new regime, see our article here).

However, the ACCC has also highlighted the prevalence of ‘strategic partnerships’ in the industry that could pose risks for competition, but which may not be captured by the new merger control regime.  These might involve collaboration between companies at different levels of the supply chain to produce a product or service on the basis of mutual exclusivity, or large-scale ‘acqui-hires’ of staff and IP from a nascent competitor to effectively acquire it without triggering merger scrutiny. See ACCC AI industry snapshot, section 4.2. This is an issue confronting regulators around the world, and the ACCC has signalled that it will be closely monitoring international developments to shape its response at home.

We can expect that, in the meantime, the ACCC will focus on how its existing enforcement tools can be used to ensure that markets for emerging technologies remain contestable.  The ACCC’s action against Google last year is a case in point.  Google admitted to contravening section 45 of the CCA by reaching an understanding with Telstra and Optus to make Google the default preinstalled search engine on Android phones. https://www.accc.gov.au/media-release/google-admits-anti-competitive-conduct-involving-google-search-in-australia. Though the contravening conduct took place from December 2019 to March 2021, Google’s enforceable undertaking to the ACCC under s 87B of the CCA, offered as part of the parties’ agreed settlement, addressed the ACCC’s broad concerns about future competitive dynamics in the industry. Google undertook to remove certain clauses from its agreements with Australian telcos that were not at issue in the proceedings, to ensure that providers of emerging search tools – including AI-enhanced tools not yet even invented – would be able to compete for preinstallation on Android phones in the future. Google’s undertaking available here. Telstra’s and Optus’s undertakings available here. TPG’s undertaking available here. ACCC media release on outcome of court proceedings available here. The case highlights how the ACCC views competition enforcement as not only about confronting the impacts of anti-competitive conduct at a given point in time, but also as a means of promoting competition in nascent and as-yet unregulated industries.

Conclusion

There is clear appetite for early intervention to address competition and consumer law risks posed by AI, and vigilance in monitoring those risks which are yet to fully manifest. The Commonwealth Government and the ACCC have adopted a strategy of articulating their concerns in real time as the technology develops. Businesses looking to deploy AI should have regard to these signals and ensure that they too are proactively identifying and mitigating these potential harms.

Contact