“We can’t hire AI talent” has become a universal enterprise complaint. But what exactly is the skills gap? Where is it most acute? And what can organisations actually do about it?

Here’s a realistic assessment of Australia’s AI skills landscape as we close out 2025.

The Shape of the Gap

The AI skills gap isn’t uniform. Different skill categories face different challenges:

ML/AI Engineering

The situation: Severe shortage. Experienced ML engineers who can take models from development to production are rare. Competition is intense.

Why it’s hard: These skills require both software engineering and ML expertise – a combination that takes years to develop. Australian universities produce graduates, but experience takes time.

Salary reality: Senior ML engineers command $200-300K+ in major markets. Organisations unwilling to pay competitive rates don’t get candidates.

Data Science

The situation: Moderate shortage, but nuanced. Entry-level data scientists are more available. Senior data scientists with business acumen and production experience remain scarce.

Why it’s hard: The “data scientist” title covers a wide range. Organisations often need specific capabilities that generic training doesn’t provide.

Salary reality: $130-200K for experienced data scientists, depending on seniority and specialisation.

AI Strategy and Governance

The situation: Emerging shortage. As organisations mature, demand for AI strategy, governance, and ethics expertise grows. Supply hasn’t caught up.

Why it’s hard: These skills combine technology understanding with business and risk expertise. Pure technologists and pure business people both lack required breadth.

Salary reality: Variable depending on role framing, but senior positions exceed $200K.

AI-Literate Business Professionals

The situation: Moderate shortage but improving. Business professionals who understand AI well enough to direct it effectively are increasingly needed.

Why it’s hard: Traditional business training didn’t include AI. Upskilling takes time and initiative.

Salary reality: Premiums of 10-20% for demonstrated AI literacy in traditional business roles.

What Organisations Are Doing

Hiring Internationally

Many organisations sponsor skilled migrants for AI roles. This works but has limitations: visa processing time, relocation complexity, retention challenges when sponsored workers gain permanent residency.

Australian immigration policy affects this lever significantly. Current settings allow skilled migration but aren’t optimised for tech roles.

University Partnerships

Partnerships with universities for graduate recruitment, research collaboration, and customised training programs. This builds pipeline but doesn’t solve immediate needs.

Universities are expanding AI programs, but graduate volume takes years to scale and graduates still need experience.

Internal Upskilling

Training existing employees in AI capabilities. This is often the most effective approach – existing employees understand the business context that external hires lack.

Effective approaches include:

• Formal training programs (internal or external)
• Project-based learning with mentorship
• Rotational assignments to AI teams
• Communities of practice for peer learning

The investment is significant but builds lasting capability.

External Partners

Using consultants and contractors to fill immediate gaps. This works for project-based needs but is expensive for sustained capability and doesn’t build internal knowledge.

AI consultants Melbourne with strong Australian presence can provide both delivery capacity and knowledge transfer, though organisations should be deliberate about building internal capability alongside external support.

Rethinking Role Requirements

Some organisations are finding success by decomposing AI roles into components that existing staff can handle:

• Data preparation: Analysts can often do this with training
• Model selection: Less skilled than model building
• AI product management: Business skills plus AI understanding
• AI operations: IT skills plus AI-specific knowledge

This approach extends limited senior AI talent further.

What’s Not Working

Competing on Salary Alone

Throwing money at the problem without addressing other factors (interesting work, growth opportunity, good management) leads to high turnover and bidding wars that benefit no one.

Expecting Juniors to Fill Senior Gaps

Hiring junior staff and expecting them to deliver senior-level work without mentorship and development time. This burns out juniors and produces poor outcomes.

Ignoring Retention

Many organisations focus on hiring while failing to retain existing AI talent. Retention is often more cost-effective than replacement.

One-Size-Fits-All Training

Generic AI training that doesn’t connect to employees’ actual work produces credentials without capability.

The University Pipeline

Australian universities have expanded AI-related programs significantly:

Computer Science with ML focus: Most major universities now offer this. Graduate quality varies by institution.

Data Science degrees: Widespread and growing. Quality increasingly good.

AI-specific programs: Emerging at several universities. Still maturing.

Short courses and certificates: Expanding rapidly. Quality varies enormously.

The pipeline is building, but it takes 3-4 years for undergraduate investment to produce graduates, plus additional years for experience development.

Government and Industry Initiatives

Several initiatives aim to address the skills gap:

Tech Skills Accelerator programs: Government-funded programs to accelerate tech skills development. Variable effectiveness.

Industry certifications: Vendor certifications (AWS, Google, Microsoft) provide structured learning paths but vary in practical value.

Professional associations: Tech industry groups offering training and networking. Useful for continuous development.

These help but don’t solve the fundamental supply constraint.

Practical Recommendations

For organisations facing AI skills shortages:

Near-term (6-12 months)

Identify critical roles. Not all AI positions are equally important. Prioritise the few that are truly critical.

Pay market rates. If you’re below market, you won’t get candidates. Accept this reality.

Use partners strategically. Fill immediate gaps with external support while building internal capability.

Retain who you have. Losing one experienced AI practitioner costs more than hiring two.

Medium-term (1-2 years)

Build upskilling programs. Identify employees with aptitude and invest in developing them.

Create attractive roles. Interesting work, good management, and growth opportunity attract candidates beyond salary.

Partner with universities. Build relationships for graduate pipeline and research collaboration.

Long-term (3+ years)

Contribute to ecosystem. Support industry initiatives, engage with education providers, share learning.

Build employer brand. Organisations known for excellent AI work attract better candidates.

Develop internal career paths. Clear progression keeps people and builds deep expertise.

Final Thought

The AI skills gap is real but not unsolvable. Organisations that combine competitive compensation, genuine development opportunity, and strategic use of external partners can build necessary capability.

The trap to avoid: waiting for the market to solve the problem. Skills shortages persist because everyone waits. The organisations that invest in building capability – through hiring, development, and retention – will outperform those waiting for easy answers.

The skills you need exist. The question is whether you’re willing to invest in acquiring and developing them.