Up and down the ASX, you’ll see companies big and small promising huge new programs of work “powered by AI”. But how can you tell if it’s legit? Here’s how to spot a company telling fibs about their AI from someone who looks into it for a living.

(For what it’s worth, “powered by AI” is my least favourite phrase in recent history.)

The hype cycle around artificial intelligence has well and truly reached the ASX. During the latest reporting season, it felt like every second company was talking up a new AI initiative. But according to Bell Direct analyst Grady Wulff, investors should tread carefully before buying into the promises.

“Be careful when you’re looking at the AI exposure this reporting season, because a lot of companies say they’re investing in it. But when you look at the financial statements, they might not have allocated any money to it,” she warned.

“So don’t get caught up with looking at the hype of the AI in the highlights of the reports. Make sure you look down into the financial sections and see what costs are allocated to AI and how much they’re going to benefit from using it.”

For investors, the lesson is simple: if the numbers don’t back the rhetoric, the AI story is likely little more than a fib. No big new dollars allocated into software, hardware or emerging tech products? It's time to start asking some questions.

According to Wulff, its not too late to get your foot in the door with AI if you're looking to ride the investment wave.

“A lot of investors have been looking for AI exposure. A lot of people think that they’ve kind of missed the boat, but they haven’t,” she said.

“We’re looking at the data centre play Goodman Group, and their expansion into data centres. That’s kind of the diversified way in which we’re playing [AI] at the moment. We’re quite bullish on Goodman Group with the buy rating. I like their expansion, what they’re doing on a global scale, because they already have the facilities for these data centres, like in Hong Kong, that it’s the logistics centres that they’re literally just putting data centres on top of, which is simply amazing.”

The Goodman example highlights the difference between companies with real, measurable AI-linked strategies and those that are simply adding the term to presentations. Wulff says the distinction is visible when you dig into the financials.

What real AI investment looks like

The difference between talk and action is clearest when you look at how some of Australia’s largest companies have approached artificial intelligence. Both Commonwealth Bank and Telstra for example have publicly committed significant capital to AI programs that are already changing the way their businesses run.

CommBank has spent years building AI into its customer service operations, fraud detection and digital banking tools. It created an in-house AI lab, has run pilot projects with global partners such as H2O.ai, and continues to scale up machine-learning systems that help personalise banking products and detect suspicious activity. Those programs appear not just in glossy strategy slides, but also in the bank’s annual results where budgets and headcount are allocated to AI development.

Telstra has also invested heavily, particularly in using AI to manage its vast telecommunications network. From predictive maintenance to AI-driven call centre assistants, Telstra’s programs have been multi-year undertakings backed by hundreds of millions of dollars in IT spend. The company has consistently flagged these projects in its results, noting the role AI plays in lowering costs and improving customer experience.

These examples illustrate what Wulff advises investors to look for: when AI is real, it is supported by funding lines in the accounts, explicit strategic programs and measurable results.

When AI is little more than a buzzword, it turns up in headlines but not in the balance sheet.