AI job titles have exploded. In Australia, Head of AI roles grew 28% in a single year, AI Engineer roles grew 37%, and somewhere in the middle of that growth, senior leaders are approving headcount for technology recruitment roles they do not fully understand.

Some AI roles are genuinely new. A decent proportion are existing jobs with "AI" bolted on the front. Five years ago it was Digital Marketing Managers, then Growth Marketing Managers, now it is AI Growth Managers. The fundamentals have not changed much. What has changed is the expectation that AI is part of the toolkit, not a nice to have.

The businesses getting this right are not hiring AI for the sake of it. They are hiring people who use AI to solve commercial problems: reduce costs, improve customer experience, speed up decisions, drive revenue. That is where the value sits, regardless of what the title says.

The questions leaders aren't asking

Too many hiring asks start with "we need an AI person." That is like saying "we need someone who is good at Excel." It is a tool, not a strategy. The question that should come first: what business problem are we actually trying to solve? Closely followed by: does our data even support an AI strategy?

From there, the checklist is short but revealing:

• What outcome do we expect this person to deliver?

• How will we measure success in six or twelve months?

• Could AI augment the existing team instead of adding headcount?

• Does this person understand our business, or just the technology?

• Can they influence and tell a story well enough to get buy-in?

The best AI hires are connectors and translators. They can speak to engineers in the morning, commercial teams in the afternoon, and executives by the end of the day. That is rarer than knowing the latest model or platform, and it is what separates a hire that pays for itself from one that quietly stalls a project.

The cost of a bad hire

The salary is usually the cheapest part. A poor senior AI hire can cost two to three times that once recruitment fees, onboarding, lost productivity, delayed projects and opportunity cost are added on. This is where the stakes are highest for financial services, government and enterprise teams already under scrutiny for how they spend AI budget.

The bigger cost is confidence. Some organisations spend twelve months recovering from a hire who over-promised what AI could deliver. The business turns sceptical, funding dries up, and genuinely valuable initiatives get shelved because everyone concludes AI did not work.AI did not fail. The hiring decision did.

What to look out for

Do not hire for AI expertise alone. Hire for commercial impact, curiosity and influencing skills. The strongest candidates do not spend the interview talking about models and prompts, they tell you how they cut costs by 30%, took a process from days to minutes, or generated new revenue.

Technology changes every six months. The ability to solve business problems does not. In twelve months there will be another wave of AI tools. The person who knows how to create value will still be valuable. The person who only knows one tool is on a slippery slope to nowhere.

Not sure whether your next AI hire is the right one

The Omnia Collective works with technology and go-to-market leaders across financial services, government and enterprise to define what an AI hire actually needs to deliver before the role goes to market, not after a bad one has already cost you a year. See our technology recruitment services or get in touch.

This post is adapted from Episode 8 of our Nap Stack podcast on AI, data, and building a business. Listen on Spotify or Apple Podcasts.