AI in the Workplace: Everyone’s Talking About It, But What’s Actually Happening?
SixthSense | August 2026
I’ve had some really interesting conversations with clients lately about AI.
Almost everyone is talking about it. Executives are asking what their organisations are doing with AI. Teams are being encouraged to explore it. Resources are being allocated and there’s a growing expectation that businesses should be doing something.
But when I ask what has actually been implemented across the organisation, the answer is often, not that much. At least not yet.
Some organisations allow ChatGPT, while others don’t allow it at all. Microsoft Copilot is probably the tool I hear mentioned most, but much of the use still seems to be around administrative tasks, drafting, summarising, meeting notes and generally helping people work a little faster.
Technology teams are also using tools such as GitHub Copilot for software development, and that opens up a different conversation again.
AI Whack-A-Mole
One client recently described the current environment as “AI Whack-A-Mole”, and I thought it summed things up pretty well.
Rather than implementing AI across an entire organisation, businesses are increasingly looking at individual problems or processes and asking whether AI can help solve them.
That makes sense.
But solve one problem here, automate another process there, introduce another AI tool somewhere else, and before long a large organisation can have pockets of AI solutions appearing everywhere.
And every one of them potentially needs to be governed, secured, maintained and paid for.
Who owns it? What data can it access? Where does that data go? How secure is it? Who maintains it? What happens when the underlying technology or model changes?
And increasingly, how much does it actually cost when people start using it at scale?
From Experimentation to Implementation
A common message employees are receiving is essentially:
Pick a process or a problem and see whether AI can solve it.
It’s a sensible way to encourage experimentation. But at enterprise scale, it can also create another problem.
There is a risk of what’s sometimes called “random acts of digital.” As teams experiment independently with Copilot, ChatGPT, automation and other AI tools, organisations can quickly end up with pockets of solutions that each need to be governed, secured, funded and maintained.
It raises a bigger question for enterprise: how do you encourage AI experimentation without creating a fragmented collection of solutions across the business?
The Cost of AI Isn’t Always a Licence Fee
This was something that really caught my attention in another recent client conversation.
They told me costs associated with their AI technology usage had gone from around $10,000 to $85,000 in a single month as consumption increased.
I haven’t independently verified those numbers, but the conversation highlighted something I think many businesses are only beginning to grapple with.
AI isn’t always a predictable per-user software licence.
Consumption-based models are becoming part of the equation. GitHub Copilot, for example, now incorporates usage-based AI Credits for Business and Enterprise customers, with consumption influenced by the models used and the amount and complexity of AI activity.
A simple AI interaction and an AI agent analysing a codebase, working across multiple files, generating and reviewing code and performing a series of tasks are very different things from a consumption perspective.
For enterprise, this potentially introduces another challenge alongside security, governance and data.
What happens to the cost when AI adoption actually succeeds?
Start With the Problem, Not the AI
Interestingly, one of the best examples of AI and automation delivering a measurable business outcome that I’ve heard recently didn’t come from a large enterprise.
It came from a small cleaning business.
I met with a consultant who had worked with the owners to understand why the business wasn’t making the money it should have been.
They didn’t start by asking, “Where can we use AI?”
They started with the problems.
Incorrect job costings were addressed. Timesheets were automated. Staff availability and scheduling were improved. Manual processes were redesigned and better information became available to run the business.
The result was remarkable.
I was told the improvement in profitability realised in the first full month of the trial, if sustained, represented an increase of around 400 per cent on the business’s previous annual profit.
It wasn’t AI for the sake of AI.
It was technology being applied to real business problems with a measurable outcome.
Maybe We’re Asking the Wrong Question
There’s clearly enormous potential in AI, and I’m certainly not suggesting organisations shouldn’t be investing in it.
But from the conversations I’m having, many are still trying to understand where it genuinely fits.
There’s experimentation happening. There are some great individual use cases. There’s pressure from executives to move faster. But there are also very real questions around data, security, governance, ownership, maintenance and now consumption-based costs.
Perhaps the starting question shouldn’t be:
“What are we doing with AI?”
Maybe it should be:
“What problem are we trying to solve?”
And then:
“Is AI actually the best way to solve it?”
Because implementing AI is one thing.
Implementing it in a way that genuinely improves the business, can be governed and maintained, protects the organisation’s data and still makes financial sense when hundreds or thousands of people start using it, is something else entirely.
About the Author
Anita Pages-Oliver is the Founder and Managing Director of Sense Recruitment and has spent almost 30 years working in recruitment across Western Australia. Through long-standing relationships with business and technology leaders across WA, Anita and the Sense team have a front-row seat to the challenges, changes and conversations shaping the workplace.
#SixthSense shares the observations, trends and practical advice we’ve gathered from almost two decades of recruiting across Western Australia. They’re the conversations we’re having every day with employers, candidates and contractors, shared to help others make better hiring and career decisions.
Sense Recruitment
Suite 6, 996 Hay Street, Perth WA
Phone: +61 8 9481 8111
Email: [email protected]
Website: www.senserecruitment.com.au
