The AI displacement risk facing admin roles – and how Australian EAs are getting ahead of it

A recent Fortune article laid out an uncomfortable statistic for EAs. The number of Americans working as secretaries and administrative assistants has fallen from 3.5 million in 2004 to 2.1 million today

America’s secretaries and administrative assistants are down from 3.5 million to 2.1 million in two decades. Two Australian EAs explain what’s kept them ahead of it.

A recent Fortune article laid out an uncomfortable statistic for EAs. The number of Americans working as secretaries and administrative assistants has fallen from 3.5 million in 2004 to 2.1 million today – and the US Bureau of Labor Statistics expects the decline to continue.

Clerical and admin roles, the article argues, may be more exposed to AI-driven displacement than most professions.

It’s a US data set but the underlying question isn’t geographically contained. As AI takes on more of the work that used to define the role, who ends up ahead of that shift and who gets left behind by it?

We asked two Australian EAs (who featured in our earlier piece on whether AI is turning EAs into something ‘more executive’). Their answers point less to displacement and more to what deliberate adaptation looks like in practice.

Process first, technology second
For Yvette Simpson, the shift hasn’t been about AI replacing tasks so much as changing how she approaches them. “AI has changed my role dramatically – sometimes in a good way, sometimes not,” she says.

She’s using it for everything from quick queries to higher-level thinking, like drafting meeting minutes within secure, locked-down programs and building planner boards with deadlines, research and recommendations. “I’m doing the same things faster, but also smarter.”

But she’s clear about the cost of staying current. “The exhausting thing can be keeping up with what it’s able to do. There’s not always time to play or learn, which means I sometimes default to ‘it’s just quicker if I do it the way I’ve always done it.'”

Her operating principle is deliberately old-fashioned in a very useful way: “I believe in process first, technology second. You can’t add the technology unless you really understand what the process is, what you want it to do and look like, and what you want the outcome to be.

“Only then can you add in the technology so you’re clear on what you’re asking it.”

Blending capability with judgement
Marianna Masullo, a fractional operations partner, describes a similar expansion rather than a straightforward speed-up. “It isn’t just about doing the same tasks faster – it’s about expanding our toolkit.

“We’re increasingly expected to blend technical AI capabilities with core human skills. While AI handles data processing and routine drafting, the real value comes from applying critical thinking, emotional intelligence and context to those outputs.”

She sees the pressure and the opportunity as inseparable. “It naturally brings more pressure because the expectations are higher but it also opens up space for greater impact.

“To thrive, we have to stay ahead of the curve through continuous learning and formal training. As the heavy lifting of routine tasks shifts to AI, our ability to connect, communicate and solve complex human problems becomes our most critical asset.”

Where that leaves the ‘more executive’ question
Neither Yvette nor Marianna frame what’s happening as the role becoming ‘more executive’ in a simple sense.

For Yvette, it’s better described as becoming more of a partner; someone who sees the bigger picture alongside their executive but with a different lens on staff engagement and internal operations.

Marianna sees it as an enhancement rather than a replacement of the assistant side of the role. The strategic footprint is growing but the foundation of trust and relationships hasn’t changed.

What both agree on is that the AI conversation isn’t optional anymore. The US data paints a stark picture of a profession under pressure.

Yvette and Marianna describe what sits on the other side of that pressure: ongoing adaptation. Learning the technology deliberately, understanding process before adding tools, and treating judgement and human connection as the parts of the job AI still can’t touch.