Apples, Oranges and Everything in Between: Are these ASIO powers really the same?
A line that keeps coming up in support of the Australian Security Intelligence Organisation Amendment Bill (No. 2) 2025 is that ASIO’s compulsory questioning powers have existed for 23 years without abuse.
It’s an effective line. It suggests a system that has been used carefully and hasn’t drifted beyond its intent. The historical record broadly supports that view. But the claim relies on a comparison that doesn’t hold up once you look at how the framework has changed.
The powers in place today are not the same as those introduced in 2003.
When compulsory questioning powers were introduced in 2003 under the Australian Security Intelligence Organisation Legislation Amendment (Terrorism) Act 2003, they were treated as exceptional.
ASIO had always been an intelligence-gathering agency. Compulsory questioning moved it into coercive territory, and Parliament responded by placing clear limits around how those powers could be used.
They were tied directly to terrorism. They were time-limited. And they were subject to a sunset clause that required Parliament to revisit them on a regular basis.
Over the past two decades, those renewal points have not been procedural. Each extension required justification, scrutiny, and political agreement.
The “23 years” being referenced sits within that structure.
The Bill now before Parliament reshapes that structure in a few important ways.
Scope
The original powers were anchored to terrorism. The proposed framework extends them into a wider set of national security categories, including sabotage, foreign interference, politically motivated violence, and threats to defence systems and territorial integrity.
These are broader and, in some cases, less tightly defined. That changes the range of situations where compulsory questioning could be considered.
Warrant process
There is also a shift in how warrants are issued.
Under the current model, the Attorney-General approves the application and a judge formally issues the warrant. That creates a clear separation between the executive seeking the power and the judiciary authorising it.
Under the proposed model, the Attorney-General would both approve and issue the warrant. A judicial figure remains involved, but not as the final decision-maker.
This adjusts where final authority sits.
Duration
The sunset clause would be removed. Instead of requiring periodic renewal, the powers would become ongoing.
If passed, the Bill would place final warrant authority with the Attorney-General of the day.
At present, that role is held by Michelle Rowland. In previous portfolios, she has overseen or supported legislation expanding regulatory oversight in areas such as online safety, misinformation, and digital platform monitoring.
The point here isn’t about an individual. It’s about direction and structure. Legislative frameworks like this carry forward across governments.
The design needs to account for how the powers could be used over time, not just under current leadership.
Another detail often left out of the broader discussion is how infrequently these powers have been used.
Across more than 20 years, there have been roughly 20 warrants issued. Most of that activity occurred earlier on, with very limited use in recent years.
That record reflects a cautious approach. It also reflects the narrower scope and the layered safeguards built into the original framework.
The argument for making these powers permanent leans heavily on their past use.
But the earlier framework and the proposed one are not the same.
The earlier model operated with:
a terrorism-specific threshold
judicial issuance of warrants
a built-in requirement for periodic parliamentary review
The proposed model would operate with:
broader national security categories
warrant authority consolidated within the executive
no expiry or mandated review point
The shift is structural.
Saying the powers have not been abused over 23 years is accurate within the conditions that applied over those 23 years.
Those conditions included tighter definitions, independent warrant issuance, and repeated parliamentary oversight.
The proposal now on the table removes or reshapes each of those elements while relying on the same track record as justification.
That leaves a gap between the evidence being cited and the model being proposed.
There is broad agreement that ASIO needs effective tools to respond to genuine national security threats.
The issue here is how those tools are designed and maintained over time.
The key question is straightforward:
Should expanded compulsory questioning powers be made permanent when the expanded version of those powers has not operated in a permanent form?
The inclusion of a sunset clause in 2003 ensured these powers would be revisited and re-examined. It kept Parliament actively engaged in deciding whether they should continue.
Removing that mechanism, while also widening scope and adjusting the warrant process, marks a clear shift in approach.
It moves the framework from one that required ongoing justification to one that does not.
That shift should be assessed on its own terms.