Defence acquisition is the largest discretionary capital program the Commonwealth runs, and the one with the longest history of schedule slip. The Major Projects Report exists to let Parliament see, project by project, whether that is still happening.

On cost it still does. The 2024-25 report carries per-project expenditure against budget for 21 major projects, with variances explained -- an overspend of $227.1 million on LAND907 Armoured Combat against a $504.0 million forecast, a 77.7 per cent underspend on LAND19 Phase 7B, and so on. That is real project-level reporting and the register should say so.

On schedule it has been hollowed out. Defence marks selected information "not for publication" following a security review, and in the 2024-25 report 19 of the 21 projects carry NFP material. Only three projects disclosed a Final Operational Capability date. Sixteen withheld it. Two others had no approved or defined FOC milestone.

The consequence is stated by the Auditor-General rather than inferred. The report records that NFP marking "has significantly increased from 2021-22 to 2024-25, reducing the ANAO's capacity to publish longitudinal analysis", and that because key schedule information including FOC is not disclosed, "analysis of the underlying basis for individual schedule slippage is no longer able to be provided", leaving only aggregate, average and median values. The total slippage across the 21 projects is 404 months. Which projects it belongs to is largely not published.

Four successive Emphases of Matter, from 2021-22 to 2024-25, have drawn attention to this in the Auditor-General's independent assurance reports. That is an auditor saying, four years running, that it can no longer tell Parliament what it used to be able to tell Parliament.

Read the Major Projects Report. Project Data Summary Sheets carry cost, scope and narrative per project, and Part 3 lists exactly which PDSS sections each project has marked NFP -- so the report is transparent about the shape of its own opacity, which is more than most withholding regimes manage.

Aggregate and median slippage are published. Attributing slippage to particular projects is possible only for the handful that disclosed FOC.

Marked verified on 2026-09-03: the report was downloaded and searched. The quoted findings sit at paragraphs 2.13 to 2.16 and 4.19 -- NFP marking "significantly increased from 2021-22 to 2024-25", and analysis of individual schedule slippage "no longer able to be provided".

Checked against the Major Projects Report on 2026-09-01. The check confirmed the entry's substance and corrected its scope.

The earlier version said project-level cost and schedule had both been withdrawn. Cost has not been: it is published per project, against budget, with variance explanations. The withholding is specific to schedule, and within schedule mostly to FOC dates. The question has been narrowed to match, because a register that overstates what is missing gets discounted on the entries where it is right.

exempt is recorded alongside the publishing blockers and is treated as partly genuine, which is why this entry is not scored as fixable without new collection. NFP marking follows a security review and representations by the Vice Chief of the Defence Force, and some capability milestones carry real classification consequences.

Two things keep the entry critical despite that. The volume of NFP marking has risen sharply since 2021-22 under an unchanged classification framework, which is a change in practice rather than in the security environment. And the ANAO retains full access to the underlying material, so the loss is public visibility rather than audit -- which means the question of what the public may see is being answered by the entity being audited.

Fairness the register should keep attached: the non-disclosure practice began in 2021-22, under the previous government, and has been maintained rather than initiated since. This entry is about a standing practice.

Companion to How much has Australia paid the United States and United Kingdom under AUKUS so far, and when?, which records a different failure in the same portfolio: money reported only as an undifferentiated program total.