APS Skills Shortages in 2026: Where the Real Gaps Are


The Australian Public Service skills shortage discussion has become so generalised that it’s lost most of its diagnostic value. Almost every agency reports difficulty recruiting; almost every workforce strategy refers to skills gaps; the APS Commission’s annual reporting confirms the broad picture. The interesting question is where the genuine pinch points are and what actually works to address them.

The 2025 workforce data and the operational experience of HR leaders across the major agencies provide a clearer picture than the headline rhetoric suggests.

The deepest gaps are technical and specialist

The genuine recruitment difficulty in 2026 sits in a relatively narrow set of specialist roles. Cyber security technical roles - particularly the senior incident response, threat intelligence, and security architecture positions - remain very hard to fill. The salary gap to private sector equivalents is significant, and the supply of suitable candidates is limited.

Cloud platform engineers and senior software architects are similar. The agencies that have made progress here have generally done so by building genuine career propositions for technical staff rather than treating them as generic APS6 or EL1 positions with a technical flavour. The agencies that haven’t restructured their technical career architecture continue to struggle.

Data scientists and machine learning engineers are in short supply across most agencies that need them. The variation in what these titles mean across organisations doesn’t help - the role advertised at one agency may bear little resemblance to the role with the same title at another.

Senior project and programme managers with experience delivering complex digital transformations are scarce. This category overlaps with the technical talent question but is distinct - the people who can land a major government technology programme are a small population and they’re consistently in demand.

The middle layer is competitive but not impossible

The middle-layer professional roles - mid-career policy analysts, project officers, business analysts, programme delivery staff - are competitive but workable for most agencies. Recruitment timelines have lengthened, the candidate pool has thinned somewhat compared to pre-pandemic norms, and salary expectations have crept up. But agencies that operate competently in the recruitment market can fill these positions on reasonable timelines.

The challenge in this segment is more about retention than initial recruitment. The internal mobility patterns within and between agencies are higher than they were five years ago, and the cost of constant churn at the middle layer adds up across the system.

Generalist roles are easier than the rhetoric suggests

The graduate intake remains over-subscribed at almost every agency. The general entry-level recruitment markets are healthy, and the APS continues to be an attractive employer for new graduates. The challenge isn’t getting people in; it’s developing them effectively and keeping them through their first decade.

Generalist mid-career APS recruitment is also less constrained than the headline picture suggests. Agencies that articulate clear value propositions and run efficient recruitment processes can fill most generalist positions on reasonable timelines.

What’s working to address the technical shortages

A few approaches have produced genuine results in addressing the technical talent gap. The senior technical specialist career pathways that several agencies have introduced have helped. These create promotion paths for deep technical staff that don’t require moving into management roles to access higher pay bands. The implementation has been uneven across agencies, but where it’s been done well, it has helped retain technical talent that would otherwise have moved.

The flexible working arrangements have been more important than some commentators initially expected. The agencies that have taken a flexible approach to hybrid working, geographic location, and working-pattern flexibility have demonstrably better access to specialist technical talent than agencies with stricter return-to-office policies. This is particularly true for the cyber and senior technical roles where the candidate pool is genuinely national or international.

The use of secondment and partnership arrangements with industry has worked for some agencies in some contexts. The integrated team models where private sector specialists work alongside APS staff on specific programmes can be effective when the governance is right. The risk is that these arrangements substitute for rather than complement internal capability building, and several agencies have learned this the hard way.

The contractor and consultant rebalancing that the Australian Public Service Commission and the agencies have been working on since the 2023 reforms has been a complicated story. The intent of moving more capability in-house is sound. The execution has been challenging in specialist areas where the in-house talent simply isn’t available at the price the APS is prepared to pay.

The AI capability question

The AI capability question deserves a section of its own. Almost every agency now has an AI strategy of some kind, and the demand for staff who can implement AI in government contexts has surged in the past eighteen months. The supply has not kept pace.

The roles in demand here aren’t only the deep technical ML engineering positions. They include AI governance specialists who understand both the technology and the public sector regulatory environment, staff who can translate between technical AI work and policy or operational requirements, and senior leaders who can set and oversee AI programmes effectively.

Agencies have addressed this in different ways. Some have built dedicated AI capability teams in-house. Some have engaged external consultancies for specific projects while developing internal capability over time. Some have used a combination - internal teams for governance and oversight, external partners for specific implementation work. The Australian AI consulting community has grown to serve this demand, and the better-quality engagements with firms like Team400 and others working in the government sector have helped agencies move faster than they would have managed alone.

The risk with the consultancy model is well-understood - it has to come with genuine capability transfer or the agency ends up structurally dependent on external support. The agencies that have set up these engagements with explicit knowledge transfer expectations have generally done better than those that haven’t.

What the next workforce strategy cycle should address

A few priorities suggest themselves for the next round of workforce strategy work. The technical career architecture work needs to continue and to extend to more agencies. The flexible working settings need to be defended against pressure to roll them back, because they’re producing real recruitment benefits. The AI capability gap needs serious attention, including realistic assessment of which capabilities need to be in-house versus accessed through partnerships.

The graduate and early-career talent that the APS continues to attract is a genuine asset, but the development pipeline that turns talented graduates into productive senior specialists takes years and needs sustained investment. The agencies that have prioritised this consistently over the past decade are now the ones with the deeper internal capabilities. The agencies that have under-invested are the ones now scrambling for external help.

The skills shortage rhetoric will continue regardless of the actual position. The useful work is in addressing the specific shortages where they’re real, with the specific interventions that have evidence supporting them.