About SIMS
How this site is put together, where its content comes from, and what it isn't.
Jump to:
About SIMS
SIMS — the Strategic Intelligence Monitoring System — catalogues and synthesises AI governance, regulation, standards, and practice developments relevant to the Australian public sector. Sources are scraped on a regular cadence; a curated weekly digest is published every Monday morning.
It is intended to be useful for AI strategy, governance, and policy advisors across Australian Federal Government departments — the people who need to keep an eye on what's happening without reading every announcement, report, and blog post themselves.
SIMS is maintained as a solo side project by a private individual. It is not affiliated with, endorsed by, or operated on behalf of the Australian Public Service, the Digital Transformation Agency, or any other government agency. It runs on a small VPS, uses a single LLM provider, and ships in small increments. The emphasis is on signal, honesty about limits, and not wasting your time.
Sources monitored
Sources are curated for relevance to Australian public sector AI work, not for exhaustive coverage of AI news. The list is small on purpose — adding sources costs effort to maintain and increases the noise floor in the digest. Most additions come from a real question someone in the APS AI community asked.
Australian government and agencies
- ACCC – News Centre Primary
- APSC – Media Releases & Statements Primary
- Attorney-General's Dept – Consultations Primary
- Attorney-General's Dept – Publications Primary
- Australian Public Service Commission Primary
- CSIRO – News Primary
- DISR – Dept of Industry, Science & Resources Primary
- DTA – Media Releases Primary
- DTA – RSS (AI tag) Primary
- DTA – RSS (all news + blogs) Primary
- Dept of Finance – News Primary
- Digital Transformation Agency Primary
- National AI Centre Primary
International regulators and standards bodies
- EU Digital Strategy – News Primary
- NIST Information Technology RSS Primary
- NIST – AI News (topic 2753736) Primary
- OECD AI Wonk Blog Primary
Research institutes and think tanks
- AI Now Institute – Publications Commentary
- Alan Turing Institute – Blog Commentary
- Alan Turing Institute – News Commentary
- Centre for AI Safety – Blog Commentary
- Good Ancestors – AI Policy & Governance Newsletter Commentary
- HAI Stanford – News Commentary
- KJR – Insights Commentary
- MIT AI Risk Repository – Blog Commentary
- Oxford Internet Institute – News Commentary
News and commentary
- Import AI – Substack (Jack Clark) Commentary
- Let's Data Science – AI Governance Commentary
- MIT Technology Review – AI Commentary
- The Gradient – Substack Commentary
How relevance is scored
Each item is read by a language model that scores it on two dimensions before it reaches the catalogue:
- Topic centrality. Is the subject of the content materially about AI or algorithmic decision-making — including the full practitioner concern set of governance, strategy, regulation, ethics, standards, risk, procurement, capability uplift, and deployment patterns? Or is AI mentioned only in passing in something that's really about privacy, online safety, or something else?
- Relevance to Australian public sector AI work. Of the items that clear the topic-centrality bar, how directly actionable is this for an APS practitioner today? Items range from "directly actionable" through "shapes how agencies should think about AI" down to "context only" and "noise."
A few diagnostic signals are recorded alongside — whether the source is Australian, whether the title mentions an Australian framework or agency, whether AI-governance terminology appears — but they don't override the language model's judgement. They exist so the maintainer can audit when the model and the heuristics disagree.
Items are filtered into the weekly digest if they clear a relevance threshold and aren't marked irrelevant by the maintainer. The digest's editorial ordering and the "this week at a glance" introduction are reviewed before publication; the per-item enrichment is not.
About the implications
Each catalogue item carries a short list of "implications for Australian agencies" — sentences suggesting what an APS team might do with the item. These are generated by a language model trained to interpret each item in an APS context, and they are not reviewed by a human before publication on individual items.
A few things to keep in mind:
- Implications are starting points for thinking, not directives or formal advice. Read them as prompts you can dismiss.
- Each implication is tagged with an action — Monitor, Consider, Decide, or Implement — that signals urgency. Monitor means watch this; no action expected. Implement means there's something to act on now. The tags route attention but don't substitute for your own judgement.
- Verify against the original source before acting on anything consequential. The original is authoritative; SIMS is a shortcut.
- The weekly digest's "this week at a glance" introduction is editorially reviewed. Per-item summaries, key points, and implications are not.
Cadence and maintenance
Sources are scraped four times a week — Tuesday, Thursday, Saturday, and early Monday morning Australian Eastern time. The weekly digest is generated and published every Monday morning, covering the week that just finished.
SIMS is maintained as a solo side project. Expect occasional outages and content gaps. Send feedback or report broken sources via the contact details on the maintainer's profile.
What SIMS is not
- Not an official government tool. SIMS is a private side project. It is not affiliated with, endorsed by, or operated on behalf of the APS, the DTA, or any other agency. Nothing here is an official government communication.
- Not a real-time alerting service. Check it deliberately; don't expect it to push to you.
- Not authoritative. Treat it as a starting point. The original sources are the authoritative record.
- Not exhaustive. The source list is deliberately curated. Many AI news streams are intentionally excluded to keep the noise floor low.
- Not a substitute for legal, financial, or formal policy advice.
Privacy
SIMS has no user accounts and no reader authentication. There is nothing to log in to as a reader.
Any future "save" or "since I last visited" features will store data in your browser's localStorage only — nothing is sent to a server. There are no third-party analytics collecting personal data on this site.