I need to vent about something that’s been building for months. The amount of AI-washing in Australian business is getting absurd, and someone needs to call it out.

AI-washing is when a company claims to use AI when what they actually have is basic automation, simple rules-based systems, or, in the worst cases, humans doing the work behind a fancy interface. It’s the tech equivalent of greenwashing, and it’s increasingly common.

What I Keep Seeing

In the past six months, I’ve encountered the following from Australian companies claiming to use AI.

A “AI-powered customer insights platform” that turned out to be a dashboard with hardcoded business rules. If your “AI” is a series of if-then statements that a developer wrote, it’s not AI. It’s software. Calling it AI is misleading.

An “AI recruitment tool” that keyword-matched resumes against job descriptions. Pattern matching against a keyword list has been around since the 1990s. Adding a modern UI doesn’t make it artificial intelligence.

A “conversational AI assistant” that was actually a decision tree chatbot with pre-written responses for every pathway. There’s no intelligence, artificial or otherwise, in following a script. These chatbots can be useful, but calling them AI misrepresents what they do.

A startup that raised $3 million on the promise of an “AI engine” that, according to a former employee, was largely powered by manual operations from a team in the Philippines. That’s not AI-assisted. That’s outsourcing with a marketing veneer.

Why Companies Do This

The incentives are obvious. AI is a magic word for fundraising. Australian VCs have explicit AI mandates. Government grants increasingly favour AI projects. Customers assume AI-powered products are more sophisticated. Job candidates want to work on AI.

When calling something “AI” gets you more investment, more customers, more grants, and more talent, the temptation to stretch the definition is powerful.

And the definition is genuinely fuzzy. There’s no legal standard for what constitutes AI in Australia. A company using basic statistical regression could technically claim to use AI, since regression is a foundational machine learning technique. The question is whether the claim creates a misleading impression, and increasingly, it does.

Why It’s Harmful

AI-washing harms everyone.

It harms customers who make purchasing decisions based on false claims about a product’s capabilities. If you buy an “AI-powered” tool expecting sophisticated intelligence and get basic automation, you’ve been deceived.

It harms investors who allocate capital based on AI claims that don’t reflect technical reality. When the next AI investment downturn hits (and it will), AI-washing companies will be exposed, and legitimate AI companies will suffer from the resulting scepticism.

It harms the AI industry by inflating expectations and eroding trust. Every AI-washed product that disappoints a customer makes it harder for genuine AI products to earn adoption. The boy who cried wolf, except it’s an industry crying AI.

It harms employees who join companies expecting to work on real AI and discover they’re maintaining basic CRUD applications with an AI label.

ASIC Should Pay Attention

In financial markets, there are rules about misleading conduct. If a company tells investors it has AI capabilities it doesn’t actually possess, that’s potentially misleading or deceptive conduct under the Corporations Act.

ASIC has been active on greenwashing, bringing actions against companies making unsubstantiated environmental claims. AI-washing is the same problem in a different domain. An ASX-listed company claiming to have proprietary AI when it’s actually using off-the-shelf tools with minimal customisation is making claims that could be materially misleading.

I’m not suggesting ASIC should regulate what counts as AI. But they should apply existing misleading conduct provisions to AI claims with the same rigour they apply to environmental claims. The precedent is already there.

The ACCC Angle

For consumer-facing products, the ACCC’s role is even clearer. If a product is marketed as “AI-powered” to consumers and that claim creates a misleading impression about the product’s capabilities, it potentially breaches Australian Consumer Law.

The ACCC has already flagged AI claims as a priority area. Good. I hope they follow through with enforcement actions against the most egregious cases, because voluntary compliance clearly isn’t working.

How to Spot AI-Washing

Here are the red flags I look for.

Vague claims with no specifics. “Powered by proprietary AI technology” without any detail about what the AI does, what models it uses, or what problems it solves.

AI as a feature label, not a capability description. “AI-powered analytics” that’s actually a standard BI dashboard. The AI label is marketing, not engineering.

Reluctance to demonstrate. Companies that won’t show the AI working in real-time on unfamiliar data. If the demo only works with pre-prepared inputs, be suspicious.

No AI talent on the team. Check LinkedIn. If a company claiming to build AI products doesn’t have machine learning engineers, data scientists, or research scientists on staff, ask who’s building the AI.

What Should Change

Australian businesses should be honest about their technology. If you’ve built good automation, call it automation. If you’ve built a rules-based system, call it a rules-based system. Both can be valuable without being AI.

If you genuinely use AI, be specific about what it does, what it doesn’t do, and what its limitations are. Customers and investors respect honesty more than hype, and the ones who don’t aren’t customers or investors you want.

The AI-washing bubble will pop. When it does, the companies that were honest about their technology will retain trust. The ones that weren’t will lose it, along with their customers and their credibility.

Don’t be the company that gets caught when the tide goes out.