‘AI workforce reset’: Why job losses don’t tell the full story for HR leaders

HR leaders are staring down a paradox. On one hand, headlines scream about large‑scale redundancies at Atlassian, Meta, Amazon and other tech giants, often framed as fallout from a rapid shift to artificial intelligence. On the other, leading experts argue that AI is more likely to reshape work than erase it – and that organisations which respond with panic cuts rather than deliberate redesign risk long‑term damage to capability, trust and competitiveness.

Speaking with HRD, Thomas Mackenzie, director of client services at talent solutions firm Scale by Avec, and Neal Woolrich, advisor in Gartner’s Human Resources group, both stressed that AI is catalysing a deep workforce reset – but not in the simplistic “robots replace humans” way many employees fear.

Redundancies are real – but AI is often the “convenient story”

From the employee vantage point, the optics are grim. High‑profile tech names are trimming thousands of roles while ramping up AI investment. It’s little wonder staff are asking if they’re next.

Mackenzie argued that while AI is absolutely accelerating change, it’s not the sole villain it’s often made out to be.

“If you ever just read the headline and not dig deeper, then yes, you're going to have a lot of alarm bells going off,” he said. “These companies… have moved from a growth‑at‑all‑costs model… where they're at now is they need to drive efficiency. One lever you can pull is drop your spend on headcount.”

He described AI as a “nice, convenient story” some organisations use to dress up broader restructuring: shifting operating models, correcting post‑COVID over‑hiring and responding to economic and geopolitical shocks. Layoffs, he points out, “have not just started in the last year – they've been going on since 2022.”

Woolrich recognised something similar in the data. Gartner’s economic modelling found that in the past year, fewer than 1% of job losses could be directly attributed to AI productivity gains. In most organisations announcing cuts, other forces – interest rates, cost pressures, market slowdowns, investor expectations and global instability – are doing at least as much of the work as algorithms.

“Quite often, AI is a secondary or even lower‑ranking factor that’s causing them to cut headcount,” he said.

This isn’t the first technology shock – and it won’t be the last

Both experts draw a straight line from today’s AI disruption back through the computer revolution of the late 20th century and the industrial revolution before that. Each period was accompanied by dire predictions of mass unemployment. Each ultimately saw employment grow – but not without painful, uneven transitions.

“We’ve seen this play out before,” said Mackenzie. “You see closed coal mines across the middle of North America, but that workforce has been redeployed. People had to learn different skill sets.”

Woolrich agreed, noting that Gartner’s modelling suggests AI is likely to be a net job creator from around 2030 onwards, even if the near‑term story feels very different.

“Every time there's been a major technological shift, employment has grown over time,” Woolrich said. “A lot of people are saying this time will be different with AI, but we're a bit more optimistic than that scenario.”

For HR leaders, that historical parallel matters. It suggests the core challenge is not preventing change, but managing redeployment, reskilling and redesign fast enough to keep people – and the business – in front of the curve.

From experience‑based hiring to skills‑based, flexible workforces

One of the clearest shifts Mackenzie is seeing is a decisive move away from hiring based on brand‑name employers – “the logo on the CV” – toward genuinely understanding what people can do and where those skills can be redeployed.

“When I first got started in recruitment, you looked at people’s logos and you go, ‘Okay, they worked at Salesforce, they can go work at Oracle,’” he said. “There are a lot of people that have worked in different industries that can move and redeploy. I think [AI] will accelerate that. So not a bad thing.”

Citing World Economic Forum projections, he noted that around 22% of jobs are expected to be disrupted by 2030, and roughly 40% of skills in those jobs will change. For HR leaders, this should sharpen the focus on:

• Real skills taxonomies, not job titles or industry labels

• Internal mobility programs that allow people to move from declining tasks into adjacent, growing ones

• Assessment methods that look at potential and transferable capabilities rather than narrow experience

At the organisational design level, Mackenzie is urging clients to abandon brittle, “all‑permanent” models in favour of more flexible, fractional and blended workforces.

“We’re looking at fractional teams more – 0.2 or 0.4 of a role versus the whole thing,” he explained. “That might be a project manager two days a week on this [project], but we can redeploy them on this other project over here. It ends up still being a full‑time resource, but they’re being deployed in different ways.”

His preferred analogy is floorboards in a house: pack them too tightly and they crack under heat; build in room to flex and they expand and contract without breaking.

“Going in and designing your whole workforce with permanent headcount may have worked in the 1980s,” he said. “We’re seeing so much shift… we need to build our workforces able to adapt so we don't have this massive upswing and downswing every time something happens.”

The new high‑value skills: Soft, human and AI‑augmented

If some work will be automated, what becomes more valuable?

For Mackenzie, AI’s immediate impact is to strip out low‑value, cognitive “busy work” and put a premium on human interaction, context and judgement.

“What AI is really bad at is making human decisions,” he said. “It struggles with context and being able to take emotion into account.”

Across technical roles, he’s seeing demand tilt away from lone coders “under the stairs” and toward engineers and developers who can articulate, collaborate and lead multi‑disciplinary projects with AI as a teammate.

“Our customers are coming to us: ‘We want more soft skills in these technical roles, because they need to be able to articulate across broader teams,’” he says.

“Dabbling” with AI isn’t enough: Build AI value creators

Woolrich’s warning to HR is blunt: there’s a chasm between employees playing with AI tools and employees actually creating value with them.

“A lot of what we're seeing at the moment is employees are just being left to dabble with AI, and it's not really creating much value,” he said. “Employees need to understand where are the pain points in their daily work that they can integrate AI and get it to add more value, make them more efficient, or generate better outcomes for customers [and] stakeholders.”

That won’t happen by accident. He argues employers must:

• Identify clear, business‑critical use cases for AI

• Provide targeted training aligned to those use cases

• Set expectations that using AI effectively is part of core performance, not an optional extra

In other words, organisations need to turn employees into AI value creators, not just AI users.

The trust gap: Job security, ethics and psychosocial risk

Layered over the skills and design questions is a more human, and legally fraught, issue: trust.

“First and foremost, [employees are asking] will I have a job in six months, 12 months’ time?” Woolrich said. “But secondly… do employees trust their employer to train them on artificial intelligence? Do they trust them to use AI ethically?”

Woolrich’s advice is to build psychological safety in two specific directions:

1. Safety to experiment – giving employees permission, tools and time to test AI in their own workflows without fear of punishment if something doesn’t work perfectly.

2. Safety to challenge – ensuring people can speak up about poor processes, unfair AI outcomes or misaligned use cases, and suggest where AI could add value, without being ignored or penalised.

Alongside that cultural work, he points to a newer frontier for change leaders: helping employees regulate emotions in the face of constant disruption.

“In the past, we used to think it was a bit intrusive to ask about employees’ emotional reactions,” he said, and many managers felt unskilled to have these conversations. “But now it's just an area we can't avoid… If we sweep these emotions under the carpet, they'll just fester and we'll end up having to deal with a bigger problem further down the track.”

What HR leaders should do now

For HR leaders grappling with AI‑linked redundancies and an anxious workforce, Mackenzie and Woolrich’s perspectives point to a clear reframing:

• Interrogate the narrative. Don’t hide broad restructuring or economic pressures behind “AI made us do it”. Employees can see through it, and trust will suffer. Be transparent about all the drivers of change.

• Design for flex, not churn. Use contingent, fractional and project‑based models to create workforce “give” rather than relying on repeated mass layoffs and re‑hiring cycles.

• Move fast on skills. Shift from experience‑based to skills‑based hiring and mobility. Map transferable capabilities, stand up internal reskilling pathways, and explicitly define the human skills that rise in value as AI spreads.

• Invest in AI literacy with purpose. Move beyond dabbling to structured, outcome‑driven AI adoption that ties directly to business pain points and performance expectations.

• Centre psychological safety and emotion. Build cultures where people can experiment and challenge, and equip leaders to have honest, emotionally intelligent conversations about AI, uncertainty and job security.

As Woolrich put it, the rise of AI is likely the biggest, broadest change to work in a generation – even larger in scope than the COVID‑era shift to remote and hybrid models. For Mackenzie, that only reinforces the need for HR to lead with clear eyes and a learning mindset.

“If you bury your head in the sand and go, ‘I've done all this study and I refuse to do any more,’ then you're really painting yourself into a corner,” Mackenzie said. “We're all lifelong learners… you’ve got to keep that mindset.”

For organisations willing to treat AI as a catalyst for smarter workforce design – rather than a blunt cost‑cutting excuse – the so‑called “AI workforce reset” may prove less a threat than a generational opportunity.