'BrAInwashing': It's a thing

5 minute read  23.06.2023 Paul Schoff, Peter Tryfonopoulos

You would have heard of 'greenwashing', but have you heard of 'brAInwashing'? We discuss the application of the Australian Consumer Law to the promotion of products and services which use AI.

Key takeouts

  • The use and promotion of AI in products or services raises similarly acute risks to those that Greenwashing raises for commercial corporations and financial institutions.
  • Under the Australian Consumer Law (ACL), promoting artificial intelligence or your products and services which use artificial intelligence requires the same discipline as any other product or service.
  • If you are promoting AI, or products and services which utilise AI, only rely on accurate data and ensure you understand the model and whether it accounts for bias.

It seems that our news feeds every day report another 'AI breakthrough'. There is no doubt that Artificial Intelligence (AI) is hot right now. Indeed, the AI hype train has moved so fast that it has crashed into its own negative feedback loop in many places – ranging from the tech industry itself to creatives in Hollywood. The Australian government is currently considering whether further regulatory and governance mechanisms are required to adapt to the fast paced change AI brings. It recently announced measures in the 2023-24 Budget to support the responsible use of AI and consulting industry to seek advice on steps Australia can take to mitigate the potential risks of AI with a focus on governance mechanisms to ensure AI is used safely and responsibly.

Amidst all that noise, it's important for businesses to remember that, while the technology available to them has changed and the law might not yet have fully caught up (and may not for some time), there are some laws which definitely do already apply. Consumer law, in particular, remains the same. Under the Australian Consumer Law (ACL), whether you're on the cutting edge of development or an early adopter in your business, promoting artificial intelligence or your products and services which use artificial intelligence requires the same discipline as any other product or service. Your obligations haven't changed - you must continue to ensure your product or service claims are true, accurate and able to be substantiated.

By now everyone has heard about 'greenwashing'. 'Greenwashing’ is when an organisation misrepresents its sustainability-related risks, business credentials, strategies or those of its products or services. We've written a guide to assist organisations to navigate ‘greenwashing’ risks.Greenwashing has become an acute source of legal risk for commercial corporations and financial institutions.

Well, we think that the use and promotion of AI in products or services raises similarly acute risks and so we have come up with a catchy name for it: brAInwashing.

Reach of the Australian Consumer Law

It is illegal under the ACL to engage in conduct, in trade or commerce, that is misleading or deceptive, or likely to mislead or deceive. It is also illegal in connection with promoting sales to make false or misleading representations as to particular matters such as performance characteristics, uses and benefits of the products or services.

Heavy penalties and other sanctions can apply to non-compliant promotion, marketing and advertising. For specific false or misleading representations an individual can be liable for a penalty up to A$2.5 million. For a corporation, the maximum penalty is up to the greater of A$50 million or three times the benefit reasonably attributable to the contravention, or (if the benefit cannot be determined), 30% of adjusted turnover during the relevant period. Other potential sanctions include running corrective advertising.

Considerations when promoting AI or products and services reliant on AI

So what are some of the considerations if you are promoting AI or products and services which utilise AI? How do you ensure truth and accuracy and an ability to substantiate claims made? Businesses should consider the following:

  • Are there reasonable grounds for claims about future matters? A representation about future matters (such as anticipated functionality) is taken to be misleading under the ACL unless there are reasonable grounds for the claim. As a developer of AI, or a business advertising a product or service reliant on AI, you must be sure that you can objectively substantiate claims made by AI or the use of AI.
  • Do you have a deep understanding of the AI functionality and how it impacts the truth of claims made in marketing your products and services? Relying on artificial intelligence, whether it is a central part of your product or a feature operating in the background, may dictate an extra layer of interrogation to ensure truth, accuracy and verification. If you are adopting third party claims about the AI, you must be satisfied about the basis of those claims. In order to do so you need to account for risk in any qualifications in the operation of the AI to ensure your claims remain accurate at all times, only rely on accurate data and ensure you understand the model and whether it accounts for bias.
  • Have you overreached in claims made? You must not make claims which are false or misleading because they exaggerate the capability, accuracy, or functionality of a product or service, even if it is unintentional due to a lack of understanding about the artificial intelligence relied upon. Of course, it is also misleading to market a product or service promoting benefits due to the use of AI when it is in fact not reliant on AI at all.
  • Are you promoting a comparison with a non-AI product and the benefits of AI? As always, particular care must be taken to ensure any comparison is fair and accurate. Is the comparison on a comparable basis?
  • Are you using AI in the creation of the promotion? Generative AI (including chatbots) can be prone to factual errors in its response, which could create their own misleading claims without human input. From within our own legal field, a lawyer in the United States recently used an AI chatbot to draft their submissions to court. The chatbot invented fake cases which obviously caused some issues later in court – happily the legal system still checks these things! If AI is used in the development of promotional material, businesses must make sure to fact-check everything developed by that AI (just as it should with a human's work).

BrAInwashing to appear on ACCC's future Compliance and Enforcement Priorities?

The Federal Trade Commission in the United States is on to this. In February 2023 it noted the ambiguity about what the term 'Artificial Intelligence' even means but noted that one thing is for sure: "it's a marketing term". Noting the importance of claims made about AI, the FTC's enforcement threat was barely veiled when it said "You don't need a machine to predict what the FTC might do when those claims are unsupported."

What about Australia's consumer law enforcer, the Australian Competition & Consumer Commission? Greenwashing is top on the list of the ACCC's 2023-24 Compliance and Enforcement Priorities but it seems likely to us that brAInwashing will find its way on to that list in coming years.

If you would like to learn more about brAInwashing or other considerations when promoting AI, contact us.