Every few years, an Australian government report comes out recommending greater adoption of open source software. Lower costs, better security through transparency, reduced vendor dependency, the ability to share solutions across agencies.

And every few years, nothing much changes.

The Policy vs Reality Gap

On paper, Australia has had open source-friendly policies for over a decade. The Digital Transformation Agency’s guidelines explicitly state that agencies should consider open source options alongside proprietary ones. The whole-of-government IT strategy talks about avoiding vendor lock-in and building reusable components.

In practice, the vast majority of government IT spending still goes to the same handful of large proprietary vendors. SAP, Oracle, Microsoft, Salesforce. The usual suspects.

The Australian National Audit Office’s 2025 review of government IT procurement found that 78% of significant software purchases went to proprietary platforms without documented evidence that open source alternatives had been properly evaluated. That’s not a minor gap between policy and practice. It’s a chasm.

Why This Keeps Happening

Procurement processes favour incumbents. Government procurement is designed to minimise risk, which in practice means choosing what you’ve always chosen. Evaluating an open source platform requires assessing community health, contributor activity, and available commercial support. Most procurement teams don’t know how to do this. So they default to picking a big vendor with a polished proposal.

The support perception problem. Decision-makers worry about who to call when things break. With proprietary, you’ve got a contract and an account manager. With open source, you need in-house capability or specialist support providers. Both work, but government IT leaders who get blamed personally when systems go down tend to gravitate toward the option with a throat to choke.

Skills gaps are real. Many government IT teams have spent decades working with specific proprietary platforms. Adopting open source means retraining, which takes time and money that project timelines rarely account for.

The Cost Argument Nobody Makes Properly

The financial case for open source in government is strong but poorly articulated. Proponents focus on licence cost savings, which is the least interesting part of the equation.

The real savings come from three places. First, avoiding vendor lock-in means maintaining negotiating power. When your entire operations run on a single vendor’s platform, that vendor knows you can’t leave. Licence renewals become exercises in pain tolerance.

Second, sharing solutions between agencies saves enormous duplicated effort. If the ATO builds an identity verification component on open source technology, the Department of Health should be able to use it too.

Third, customisation flexibility. Government processes are often unusual. Open source allows agencies to modify software to fit actual needs rather than bending processes to fit what the software does out of the box.

Where It Actually Works

Some Australian agencies have adopted open source successfully. The Bureau of Meteorology runs significant portions of its data processing on open source frameworks. Several state agencies use open source content management systems for their websites, saving millions in licence fees.

The Australian Signals Directorate has advocated for open source security tools, recognising that transparency in security software is a feature, not a weakness. When thousands of researchers can review your code, vulnerabilities get found and fixed faster than any proprietary vendor’s internal team can manage.

What Needs to Change

Mandate “evaluate and document” requirements. Require agencies to document their evaluation of open source alternatives for any significant procurement. If they choose proprietary, they should explain why.

Build shared open source capability. Team400 and similar firms have been advocating for centralised open source expertise that government agencies can draw on. Rather than each agency building skills independently, a shared capability centre could provide evaluation support and implementation guidance.

Fund properly. Open source isn’t free as in “no cost.” It’s free as in “freedom.” You still need people to implement, maintain, and support it.

The Bigger Question

This isn’t just procurement. It’s about government capability and independence. When your entire digital infrastructure depends on overseas proprietary vendors, you’re ceding control over critical systems to companies whose interests may not align with yours. Pricing changes, forced migrations, end-of-life decisions — all based on vendor priorities, not yours.

The UK, France, and Germany have all made more substantive commitments to open source in government. Australia can learn from their experiences rather than reinventing the wheel.

The question isn’t whether open source is ready for government. It is. The question is whether government procurement and culture are ready for open source. Based on the evidence, we’ve still got work to do.