Item Catalogue
AI governance, regulation, strategy, and practice developments from monitored sources.
- Week of 7 July 2025
-
AI Policy and Governance Newsletter — July 2025
- July 2025 newsletter covers EU GPAI Code of Practice, US AI Action Plan, Swiss sovereign LLM, and OpenAI's Australian economic blueprint.
- OpenAI's economic blueprint for Australia and commentary on sovereign AI options are directly relevant to current APS strategy debates.
- The newsletter is a curated think-tank digest, offering useful orientation but limited depth on any single development.
- Week of 23 June 2025
-
AI is driving growth in jobs, research and innovation across Australia
- NAIC and CSIRO's 2025 AI Ecosystem Report shows AI hiring has tripled since 2015, with 1,532 organisations recruiting AI-skilled workers in 2024.
- Australia holds just 0.18% of global AI patents over the past decade, flagging a commercialisation gap that upcoming government plans aim to address.
- The report references an upcoming AI Capability Plan and Strategic Examination of R&D — both worth tracking for APS workforce and innovation policy implications.
- Week of 16 June 2025
-
Making generative AI trustworthy and reliable for adoption at scale
- Alan Turing Institute researchers reflect on their involvement in a Global AI Assurance Pilot.
- The pilot addresses trustworthiness and reliability of generative AI as prerequisites for scaled adoption.
- Extracted text is minimal - substantive findings and methodology are not available from this excerpt.
- Week of 9 June 2025
-
AI Policy and Governance Newsletter — June 2025
- Good Ancestors' June 2025 newsletter surveys AI governance developments across California, G7, Australia, and frontier AI.
- PM Albanese signed a $20B AWS data centre deal; Good Ancestors flags unanswered questions about compute access and revenue terms.
- G7 AI statement prioritises adoption over safety governance; newsletter notes minimal attention to bias, misinformation, or oversight gaps.
- Week of 2 June 2025
-
Artificial Power: 2025 Landscape Report
- AI Now Institute's 2025 report argues AI companies have accumulated disproportionate market and political power.
- Report proposes concrete strategies for policymakers to rebalance AI governance toward public accountability.
- US-centric advocacy framing limits direct applicability to Australian regulatory or procurement contexts.
-
AGI is Not Multimodal
- The article argues multimodal scaling will not achieve human-level AGI, lacking embodied reasoning.
- Proposes embodiment-first AI architectures as a more promising path to general intelligence.
- Primarily a theoretical opinion piece - limited direct operational relevance for APS AI practitioners.
- Week of 19 May 2025
-
Introducing Gambit: a tool for doing computation in game theory
- The Alan Turing Institute has released Gambit, a computational tool for game theory analysis.
- The tool targets researchers studying strategic interactions across multiple domains.
- No AI or algorithmic governance content present - this is a mathematics/research software item.
- Week of 12 May 2025
-
Taxonomy of Risks Posed by Language Models
- MIT AI Risk Repository spotlights a 2022 Google DeepMind taxonomy covering six domains and 20 subdomains of LLM risk.
- The taxonomy distinguishes observed versus anticipated risks - a useful framing for APS risk assessments and AI governance documentation.
- The underlying paper is from 2022; the blog post adds no new analysis, limiting its immediate signal value.
- Week of 5 May 2025
-
AI Policy and Governance Newsletter — May 2025
- Trump's revocation of Biden-era AI safety EOs and memoranda materially shifts US federal AI governance posture.
- Federal Court of Australia opens public consultation on generative AI use in legal proceedings, closing 13 June.
- OpenAI's 'OpenAI for Countries' initiative raises sovereignty and risk-distribution questions for potential Australian participation.
- Week of 21 April 2025
-
AI Risk Repository Report updated (April 2025)
- MIT's AI Risk Repository now covers 65 frameworks and 1,612 unique coded risk entries, updated April 2025.
- A new multi-agent risk subdomain has been added, reflecting emerging governance gaps in agentic AI systems.
- The repository supports auditing framework development and policy transparency - directly applicable to APS risk work.
-
The Risks of Machine Learning Systems
- The MLSR framework categorises ML risks into first-order (design/development) and second-order (real-world interaction) effects.
- The taxonomy integrates algorithmic impact assessments, incident reports, and ML literature - useful for structured risk assessment work.
- This is a 2022 arXiv paper spotlighted by MIT AI Risk Repository; not new guidance but a curated reference resource.
-
New Report on the National Security Risks from Weakened AI Safety Frameworks
- AI Now Institute report argues industry-led AI safety frameworks are eroding established military and critical infrastructure safety standards.
- Report calls for democratic oversight and traditional TEVV evaluation frameworks for safety-critical and military AI deployments.
- Primarily a US national security framing; limited direct applicability to Australian civilian agency AI governance contexts.
- Week of 7 April 2025
-
Mapping Frameworks at the Intersection of AI Safety and Traditional Risk Management
- MIT AI Risk Repository identified 11 frameworks bridging traditional risk management and AI safety, all from 2023 or newer.
- Frameworks draw on proven methods from aviation, nuclear, and cybersecurity to address frontier and general-purpose AI risks.
- APS agencies building AI risk governance could use this as a curated starting point, avoiding duplication of existing work.
- Week of 31 March 2025
-
Explore the Frameworks Behind the AI Risk Repository
- MIT AI Risk Repository v3 now publicly documents all 65 source frameworks behind its AI risk taxonomy.
- The resource supports policymakers and governance professionals in auditing and building AI risk classifications.
- Directly useful for APS teams developing or reviewing AI risk frameworks and governance documentation.
-
Repository Update: April 2025
- MIT AI Risk Repository Version 3 now contains over 1,600 coded AI risk categories drawn from 65 published frameworks.
- Nine new frameworks added include the final International Scientific Report on the Safety of Advanced AI and EU AI Act-aligned systemic risk taxonomies.
- A new subdomain on multi-agent risks addresses failure modes like miscoordination, conflict, and collusion in interacting AI systems.
- Week of 17 March 2025
-
Project Aardvark: reimagining AI weather prediction
- The Alan Turing Institute's Project Aardvark applies machine learning to improve weather prediction globally.
- The project targets underserved regions including the Global South and Arctic, with equity implications for climate risk.
- Limited direct relevance to APS AI governance work; primarily a research and technical development item.
- Week of 3 March 2025
-
AI research is increasingly vulnerable to state threats. Here’s how UK academia can minimise risk
- The Alan Turing Institute's CETaS calls for balancing academic freedom with AI research security against state threats.
- Australia faces analogous tensions around foreign interference in university AI research, governed by the UFIT scheme.
- Only a blog summary is available - substantive analysis requires reading the underlying report.
-
Sources of Risk of AI Systems
- A 2022 academic paper taxonomises AI risk sources across ethical and reliability-robustness dimensions.
- The MIT AI Risk Repository spotlights this as one of fourteen risk frameworks it has catalogued.
- The underlying paper is three years old; the blog post adds no new analysis beyond a brief summary.
-
Highlights from Paris: Attending the 2025 IASEAI Conference
- MIT AI Risk Repository was presented at the inaugural IASEAI conference and Paris AI Action Summit in February 2025.
- The event highlighted a fragmented international AI governance landscape, reinforcing demand for systematic risk approaches.
- This is a conference recap with limited new substantive content - low priority for APS readers seeking actionable guidance.
-
NIST Finalizes Guidelines for Evaluating ‘Differential Privacy’ Guarantees to De-Identify Data
- NIST has finalised SP 800-226, providing guidelines for evaluating differential privacy guarantees in data analytics.
- The guidelines help organisations assess vendor DP claims and navigate privacy-utility trade-offs in statistical data release.
- Primarily a data privacy standard; AI is not the subject - relevance to APS AI governance is indirect.
-
AI UK 2025 to welcome leading women from the world of AI and tech
- Alan Turing Institute's AI UK 2025 conference will feature prominent women in AI and tech.
- Item is a promotional blog timed to International Women's Day, not substantive AI analysis.
- Limited direct relevance to Australian federal agencies - included for completeness only.
- Week of 24 February 2025
-
Evaluating the Social Impact of Generative AI Systems in Systems and Society
- MIT AI Risk Repository spotlights a 2023 framework for evaluating generative AI's broad social impacts across 11 categories.
- Framework covers both technical system evaluation and societal effects, including bias, privacy, inequality, and labor impacts.
- Useful reference for APS agencies developing AI impact assessment processes, though published in 2023 and not Australia-specific.
- Week of 17 February 2025
-
AI Now Coauthors Report on Surveillance Prices and Wages
- AI Now Institute co-published a report on harms caused by algorithmic surveillance of prices and wages.
- Algorithmic price and wage surveillance has emerging relevance for Australian consumer and competition regulators.
- Extracted text is minimal; full report content and methodology are not assessable from this summary.
-
Data Strategy 2025–2027
- Attorney-General's Department released its Data Strategy 2025–2027 covering four focus areas including AI and analytics.
- AI and machine learning are referenced as tools to improve evidence-based policy in the justice sector.
- AI is a secondary thread in a broader data governance strategy; not a focused AI governance document.
-
Seven sessions not to miss at AI UK 2025
- Alan Turing Institute's Chief Scientist highlights seven sessions at the AI UK 2025 conference.
- No session content or details are available in the extracted text - only a teaser headline exists.
- Low signal for APS readers; a UK conference preview with no substantive content to evaluate.