Insights & perspectives on modern recruitment

The data landed this week and it is bracing.

AI-related job postings in the US rose 95% in the first half of 2026 compared to a year earlier, while overall job postings fell 16%. ManpowerGroup described the result as an "unbalanced" hiring landscape, with demand for AI-related positions pulling the market up while other careers come under increasing pressure.

That is not a blip. That is a structural divergence, and it is reshaping recruitment in ways that most agencies and in-house teams have not yet fully reckoned with.

Here is the uncomfortable truth: the labour market your clients think they are operating in and the labour market that actually exists are no longer the same thing. And the recruiters who keep treating them as one unified market are going to make expensive mistakes.

Two Markets, One Misleading Headline

The labour market right now is not one market. It is two parallel realities operating under the same headline, moving in opposite directions. On one side, there is explosive demand for anyone who can build, deploy, or manage AI systems. On the other, a quiet contraction in traditional roles, frozen entry-level pipelines, and a "low-hire, low-fire" environment that shows no signs of changing course.

This split is not just about AI versus non-AI roles in the abstract. It runs through individual sectors, seniority levels, and even specific functions within organisations.

PwC's 2026 Global AI Jobs Barometer, which analyses over a billion job ads from six continents, reveals that AI is creating a two-track labour market where skills like judgement and leadership are more critical and more rewarded than ever, and that companies making the biggest AI gains are actually raising wages and headcount faster than their least-AI-exposed peers.

That last point deserves to sit with you for a moment. AI adoption is not universally suppressing headcount. In certain pockets of the market, it is accelerating it. The firms that are winning are the ones leaning into AI capability, not running from it.

The Junior Talent Pipeline Is Breaking

Perhaps the most consequential, and most under-discussed, consequence of this split market is what is happening at the entry level.

The Rezi research team's 2026 report frames this as a fundamental economic shift rather than a cyclical downturn. Entry-level roles in the US are down 35% overall, with the sharpest declines in software development and data analysis where junior postings have plummeted by 67%.

If a company stops hiring junior developers because AI handles that tier of work, there is no natural pipeline producing mid-level developers five years later. The long-term consequences of this pipeline disruption could be severe, even if the short-term cost savings feel attractive.

For recruiters, this matters in two specific ways. First, clients who stopped investing in graduate and junior hiring are going to face a capability gap in three to five years that they will not be able to solve quickly. Second, the nature of entry-level work that does exist has fundamentally shifted.

The most AI-exposed junior roles are seven times more likely than the least AI-exposed junior roles to demand traditionally senior skills like leadership, and so-called "seniorised" entry-level roles are actually thriving, showing 35% growth since 2019.

Recruiters need to be reframing the conversation with clients about junior hiring. It is not about hiring fewer grads at lower cost. It is about hiring fewer grads at higher capability, earlier in their careers, with a clear development pathway that AI tools help accelerate.

What "Specialist Community" Means Now

To move with the AI boom, recruiters are rethinking their entire strategy. Firms like Hays are increasingly building expertise around specialist-skill communities rather than broad tech categories.

That is a meaningful shift in how recruitment businesses position themselves. Broad coverage is becoming less valuable. Deep contextual knowledge is becoming more so.

Demand for roles that support AI integration, including titles like AI trainer and process automation specialist, is soaring, with vacancy rates above 25% for some positions. Other sought-after roles include data scientists, database architects, and cybersecurity experts, jobs that have existed for decades but are now in even higher demand when AI skills are attached to them.

The practical implication for agency recruiters is that generalist coverage models are being stress-tested. A client looking for a machine learning engineer who can also drive cross-functional adoption does not need a recruiter who covers "tech broadly." They need someone who understands what good looks like inside that specific intersection of skills, who has a relevant network, and who can credibly advise on whether the brief is even realistic given current supply.

A generic AI model may identify broad resume keywords, but keyword overlap is not the same as fit. A legal operations hire, a nonprofit executive, a clinical leader, or a software engineer may all look qualified on paper for very different reasons. Precision requires context: the function, the seniority level, the team environment, and the employer's non-negotiables.

That context is not something an algorithm provides. It is what a specialist recruiter earns the right to offer over time.

The Mandate to Advise, Not Just Fulfil

Fewer roles are opening, but the ones that do open are harder to fill, more strategic, and more competitive.

This is precisely the environment where transactional recruiters, those focused purely on filling a brief as written, struggle, while advisory recruiters thrive. Clients with a brief for an "AI integration lead" may not yet know that what they actually need is someone with change management experience as much as technical capability. That gap between the brief and the actual need is where a great recruiter earns their margin.

The most useful mental model right now is that the recruiter's title has stayed the same but the job description has quietly rewritten itself. The recruiter is moving from executor to orchestrator: less the person who runs every search and sends every message, more the person who directs a set of tools, reviews their output, and owns the human moments that decide a hire.

That orchestrator role is not just about using AI tools more efficiently. It is about having enough market intelligence to interpret what the tools surface. When a search returns 200 profiles who technically match a job description, the value is in knowing which 12 are genuinely relevant, and why. That judgement comes from sector knowledge, from relationships, and from understanding how this specific client operates. It is human work, augmented by machine speed.

Proactive Beats Reactive in a Split Market

The same "low-hire, low-fire" environment that defined much of 2025 looks likely to continue. Overall job postings are flat or declining, but small pockets of growth are emerging as employers concentrate their limited hiring on roles and skills tied to AI.

In that context, waiting for briefs to land is a losing strategy.

The recruiters who are winning in this environment are not the ones with the most reactive capacity. They are the ones who are mapping the pockets of growth before their clients ask, who understand which roles are becoming structurally harder to fill, and who can walk into a client meeting and say: here is what we are seeing in your sector, here is where the supply problem is going to hit you in the next 12 months, and here is how we suggest you get ahead of it.

That kind of proactive, insight-led approach is exactly what Floats is built to support, giving recruiters the tools to surface relevant talent and take ideas to market before they become someone else's reactive emergency.

The Recruiter Who Reads Both Tracks Wins

The split labour market is not a temporary anomaly to be managed until conditions normalise.

Skills needed for the most AI-exposed jobs are changing more than twice as fast as for the least AI-exposed jobs, and jobs "professionalised" by AI are growing twice as fast as jobs "democratised" by AI, with 42% faster wage growth since 2021.

This bifurcation is going to deepen before it levels out.

The recruiters who will come out of this period with stronger businesses are not necessarily those who became AI experts overnight. They are the ones who understood that the market had split, updated their mental models accordingly, and repositioned their value around intelligence and judgement rather than activity and volume.

Two markets. One team. The question is which version of your market you are actually serving.