This week's most significant development for Commonwealth practitioners is the DTA's launch of a centralised AI transparency statement register on digital.gov.au, consolidating statements from all 94 reporting entities under the Policy for the Responsible Use of AI in Government alongside 20 voluntary contributors. The register is designed as a live tracking tool, and the DTA has signalled forthcoming work including an agentic AI addendum to the technical standard and a new AI Review Committee for whole-of-government oversight, with further detail expected mid-2026. Beyond the domestic picture, international material this week touches on the governance implications of AI operating across multi-agent environments and in politically sensitive contexts, themes with emerging relevance as Australian agencies consider how existing accountability frameworks apply to more autonomous AI deployments.
The DTA has launched a centralised register of AI transparency statements on digital.gov.au, consolidating publicly available statements from all 94 reporting Commonwealth entities required under the Policy for the Responsible Use of AI in Government, plus 20 voluntary contributors. The register is intended as a live tracking tool, not a static archive, with DTA providing ongoing briefings to accountable officials to improve statement quality. Alongside this, DTA flagged forthcoming work including an agentic AI addendum to the technical standard for government AI use and the establishment of an AI Review Committee for whole-of-government oversight, with details due mid-2026.
Implications
ImplementAgencies could review their published transparency statement against peers now visible in the centralised register and update content to reflect current AI use.
MonitorGovernance and policy teams could watch for the agentic AI addendum to the technical standard and AI Review Committee details, both expected mid-2026.
ConsiderAccountable officials could consider engaging with DTA's regular briefing program to support ongoing uplift of transparency statement quality.
Implications are AI-generated. Starting points, not advice.
An OECD AI blog post introduces the VIADUCT project, which examines ethical and sustainable AI training data sharing as an alternative to broad web scraping. The project addresses challenges including data scarcity despite apparent abundance, copyright issues, GDPR compliance, trust, and fairness in AI data ecosystems. The extracted content is a brief teaser only; the full post would need to be read directly for substantive analysis.
Implications
MonitorAgencies involved in AI procurement or data governance may want to monitor VIADUCT outputs as international consensus on ethical training data practices develops.
Implications are AI-generated. Starting points, not advice.
Global30 Mar 2026Import AI – Substack (Jack Clark)
This edition of Import AI covers three distinct research threads. The most policy-relevant is a Stanford political economist's argument that AI could function as 'political superintelligence' - augmenting how citizens, representatives, and institutions understand data, represent interests, and make decisions. The framework identifies three layers requiring intentional design: information access, AI delegates acting on behalf of citizens, and governance structures to prevent private AI infrastructure from capturing political agency. The newsletter also covers a robotics paper on AI-controlled drumming hands (concluding robots remain poor drummers) and a Google alignment paper on multi-agent AI society.
Implications
MonitorPolicy and governance teams may want to monitor the 'political superintelligence' discourse as it develops, given its direct implications for AI use in citizen engagement and automated government decision-support.
ConsiderAPS AI governance practitioners could consider whether the three-layer framework - information, representation, governance - offers a useful lens for assessing AI use cases in public consultation or policy development contexts.
Implications are AI-generated. Starting points, not advice.