Artificial Intelligence ('AI') is the talking point of social media, a topic of board discussion, and the key selling point in many products, and rightfully so. AI has the potential to transform every aspect of our working lives by promising disruption to our routine tasks at work.
What happens then, when organisations' jump the gun' and mislead customers by promoting their products and services as being on the 'cutting-edge of AI' when they really aren't? This question was answered by the U.S. Securities and Exchange Commission (SEC), which, in recent weeks, issued hefty fines to two investment advisory firms for "false and misleading statements about their purported use of artificial intelligence".
The conduct of these two businesses were coined as 'AI-washing' by the SEC, who stated "We find that Delphia and Global Predictions marketed to their clients and prospective clients that they were using AI in certain ways when, in fact, they were not,". The name was taken from greenwashing – where, in a similar vein, organisations falsely mislead consumers that their products are environmentally friendly.
What does this mean for Australia?
The SEC's decision is sure to place Australian regulators such as the Australian Securities & Investments Commission (ASIC) and the Australian Competition and Consumer Commission (ACCC) on notice. We believe that ASIC and ACCC will begin firming their stance on AI-washing and potentially issue fines. Organisations within Australia who use AI-buzzwords when promoting their products or services should use caution. Similarly, organisations that purchase such products should also be wary of being duped by misleading comments about a products use of AI.
How will "AI-washing" be regulated?
Australia's AI Ethics Framework provides guidance but lacks specific laws or penalties which would potentially address AI-washing. However, we can learn a good deal about how ACCC and ASIC might tackle this issue by looking at their past approach to greenwashing.
Organisations that mislead or make false claims about their AI use may be in breach of Australia's competition and consumer laws. For example, the Australian Consumer Law prohibits organisations from making false or misleading claims. Breaching these laws result in heavy civil and even criminal penalties.
The Ethical Implications
Beyond legal ramifications, AI-washing also raises substantial ethical concerns. 'AI' is an umbrella term that encapsulates various technology types – machine learning, generative AI, and natural language processing, among others. The misuse of AI-related terms contributes to confusion and misinformation for clients and investors about an organisation's capabilities. For example, it is widely known that Google uses AI in its algorithm when presenting search recommendations. However, if an organisation claims to use AI on the basis that they use Google, it would be considered misleading conduct and, accordingly, unethical.
The Australian Government's AI Ethics Framework attempts to guide organisations in the responsible design, development, and implementation of AI. However, this is insufficient in providing organisations with comprehensive guidance on how they may avoid AI-washing. This is why sellers of products and services that claim the use of AI should be cautious. One of the best things they can do is be transparent and candid with the types of technology and AI they use.
What Next?
A new and rapidly evolving area such as this can be a minefield for organisations that sell, build or procure technology products and services. Here are a few principles we implement when proposing our own AI-digital transformation products to our clients:
- Be honest and clear: Make sure to understand how AI is used in your technology before making claims. Be accurate, not hyperbolic. Clients love transparency and clarity.
- Be open about risk: Ensure you understand the risks of the specific AI in your technology. Be clear to prospective clients about the risks and how they are being mitigated. If you are procuring a product, make sure to ask for a risk strategy if you.
- Architecture designs: We provide high-level architecture designs of our products that are easy-to-understand and list each type of technology used. Request for this if you are procuring a product.
- Seek the experts: AI regulation is an ever-evolving landscape, so seek regular advice from experts in the field, this may be lawyers that deal with AI and digital tech law regularly or IT experts. They can help ensure you're keeping up with current disclosure requirements and staying within the legal bounds.
Conclusion
It is likely that the ACCC and ASIC will follow suit, putting down their foot on AI-washing. AI is the technology and innovation of our time. We should ensure it is able to prosper without being thwarted by the misleading behaviour of the minority. When buying or selling AI-integrated products, remember to opt for transparency. If anything seems fuzzy, seek advice from experts well-versed in technology.
