Item Catalogue
AI governance, regulation, strategy, and practice developments from monitored sources.
- Week of 18 May 2026
-
NIST NCCoE Genomic Data PETs Testbed & Dioptra Webinar
- NIST NCCoE is showcasing its PETs Testbed and Dioptra AI security testing platform in a June 2026 webinar.
- The testbed evaluates differential privacy and federated learning for protecting sensitive data during AI model training.
- Primarily a US-focused event; limited direct APS applicability unless agencies work with sensitive data AI pipelines.
-
Stanford HAI Launches AI and Organizations Lab to Study Science of AI in the Workplace
- Stanford HAI has launched a new lab focused on AI's real-world effects on jobs, teams, and organisational performance.
- Research outputs could inform APS workforce and AI adoption strategy as empirical evidence on AI-workplace impacts matures.
- Item is a brief launch announcement with minimal detail - substantive findings are not yet available.
-
Researchers Worldwide Compete to Shape the Future of AI in Organizations
- Over 200 academic teams submitted proposals to Stanford HAI's AI for Organizations Grand Challenge.
- Research focus is on how AI transforms teamwork and collaboration - relevant to APS workforce AI integration.
- No outputs published yet; this is a competition announcement with limited substance for APS practitioners.
- Week of 11 May 2026
-
AI Policy and Governance Newsletter — May 2026
- DTA's Policy for the Responsible Use of AI in Government v2.0 is now mandatory, with APS-wide compliance due mid-2026.
- Expert open letter urges Australia to use existing biosecurity powers to address AI-enabled bioweapons risk without new legislation.
- Australia is excluded from Anthropic's Project Glasswing defensive coalition; US is moving toward mandatory pre-release AI testing.
-
Amazon employees automate tasks with MeshClaw
- Amazon employees gamed internal AI usage metrics by automating trivial tasks to inflate token consumption scores.
- The case illustrates metric design risk: raw token counts are poor proxies for genuine AI productivity gains.
- Limited direct APS relevance, but the governance pitfalls map onto any agency deploying AI adoption KPIs.
-
AI chatbots are giving out people’s real phone numbers
- AI chatbots including Gemini and ChatGPT are exposing real personal phone numbers and home addresses from training data.
- DeleteMe reports a 400% increase in customer queries specifically referencing generative AI privacy breaches in seven months.
- Australian agencies holding PII should note this as a live risk for staff and public use of commercial AI tools.
-
New project to build trust in AI for air traffic control
- The Alan Turing Institute is building the first open-source toolkit for continuous AI trust assurance in air traffic control.
- Open-source release means Australian agencies and CASA could reference or adapt the toolkit for high-stakes AI assurance.
- Limited extracted text available; full scope and methodology of the project are not assessable from this item.
-
Building and procuring sustainable Defence AI will boost force resilience
- Alan Turing Institute research links sustainability measures in Defence AI procurement to improved force resilience.
- Findings may inform how Australian Defence and other agencies approach AI procurement risk and lifecycle planning.
- Extracted text is truncated - full substance of findings is not available for detailed assessment.
-
Experts Say Divergent Definitions Stall Global AI Governance
- An op-ed by UCL academics argues divergent AI definitions are the primary barrier to functional international governance.
- Australia sits in the middle of this dynamic - balancing alignment with major AI powers against domestic regulatory sovereignty.
- This is an opinion piece diagnosing a known problem; it announces no new rules, bodies, or binding agreements.
-
Anthropic Declines Chinese Request for Mythos Access
- A Chinese think tank representative privately asked Anthropic to grant access to its newest frontier model, Mythos, and was refused.
- US National Security Council officials were alerted and reacted with concern, signalling escalating scrutiny of informal AI access channels.
- No technical details about Mythos were disclosed; governance implications centre on access controls, not capability specifics.
-
Establishing AI and data sovereignty in the age of autonomous systems
- Enterprise survey finds 70% of global executives believe they need a sovereign AI and data platform to succeed.
- AI and data sovereignty — reducing dependence on centralised cloud providers — is an emerging policy and procurement concern for governments globally.
- This item is sponsored content from EDB via MIT Technology Review's custom content arm, not independent editorial journalism.
-
NIST NCCoE Cyber AI Profile Virtual Working Session Series: Usability of the Profile
- NIST NCCoE is running a virtual working series to refine the Cybersecurity Framework Cyber AI Profile.
- Session three focuses on usability across AI roles - users, developers, and deployers - and delivery formats.
- Open to government participants; Australian agencies could observe but this is a US-led standards process.
-
Taiwan Builds Integrated Health Data Platform for Smart Medicine
- Taiwan has established three national AI governance centres for health AI: responsible AI, external validation, and clinical impact evaluation.
- The structural split between governance, independent testing, and health-technology assessment mirrors approaches other jurisdictions are considering for AI assurance.
- Limited direct relevance to Australian federal agencies; useful as a peer-jurisdiction reference for health AI governance architecture.
-
Decisions Included in Forrester Adaptive Process Orchestration Landscape
- Forrester's Q2 2026 APO Landscape covers 35 vendors combining AI agents with deterministic and nondeterministic workflow automation.
- Forrester formalises governance, auditability, and human-in-the-loop controls as core procurement criteria for agentic automation platforms.
- This is a vendor PR announcement with analyst market framing - no new technical capability or Australian regulatory development.
-
The shock of seeing your body used in deepfake porn
- Non-consensual AI deepfake pornography causes serious reputational, financial, and psychological harm to real people.
- Australia's Online Safety Act and eSafety Commissioner have jurisdiction over non-consensual intimate image abuse including deepfakes.
- Item focuses on US adult content industry harms; limited direct APS governance or policy action implied.
-
Essay Critiques AI Use Scales' Practical Coherence
- An essay argues AI use scales in education are unenforceably vague, shifting accountability onto implementers.
- The critique parallels APS AI governance challenges: layered permission frameworks can obscure who is accountable for policy breaches.
- Source is a commentary piece, not empirical research; signal value is modest for federal agency practitioners.
-
Data readiness for agentic AI in financial services
- 57% of financial organisations are still developing internal capabilities to fully leverage agentic AI, per Forrester.
- Agentic AI use cases in financial services include regulatory reporting, trade monitoring, and client risk exposure scanning.
- This is sponsored content from Elastic via MIT Technology Review's custom content arm - not independent editorial journalism.
-
Alation launches AI Governance system of record
- Alation launched a commercial AI governance product creating a centralised inventory of models, agents, and tools with compliance workflows.
- The regulation registry references EU AI Act, NIST AI RMF, and ISO 42001 - but not Australia-specific frameworks.
- This is a vendor product announcement; APS agencies would need to assess fit against their own governance obligations and existing platforms.
-
ATxSummit Convenes Global Leaders to Shape Asia AI Agenda
- ATxSummit 2026 convenes 4,000+ leaders from 50+ countries in Singapore on 20-21 May 2026.
- Agenda themes include practical AI governance, agentic systems, and AI at national scale for public good.
- This is a conference announcement with no published outputs yet - signal value is limited until post-summit artefacts emerge.
-
SMBs Face Gaps in AI Governance and Security
- SMBs are struggling to govern rapid AI adoption, with shadow AI posing rising operational and security risks.
- Limited relevance to APS agencies directly, but patterns mirror resource-constrained teams within government.
- Item is secondary commentary aggregating other reporting - low evidential weight for APS practitioners.
-
How Chinese short dramas became AI content machines
- Chinese short drama studios now produce ~470 AI-generated episodes daily, cutting costs by up to 90%.
- Algorithmically optimised, AI-mass-produced content is reaching Australian consumers via mainstream social platforms.
- Limited direct governance relevance for APS agencies; more relevant to media regulation and online safety than AI policy.
-
Web Summit Vancouver Frames Debate Over AI Ownership
- Web Summit Vancouver drew 20,000 attendees to debate open-source versus closed-source AI governance and digital sovereignty.
- Canada's first AI Minister and PacifiCan investments signal active Canadian federal positioning on AI - a peer jurisdiction worth watching.
- This is a conference recap with no binding policy outcomes; limited direct relevance to Australian federal agencies.
-
The Download: the hantavirus outbreak and Musk v. Altman week 2
- A mixed technology news roundup covering AI workforce discontent, lawsuits, hacking, and chip supply.
- No single item reaches the depth or specificity useful for APS AI governance or policy work.
- Low signal for APS readers; AI threads are surface-level across a broad news digest format.
-
The Download: China’s AI drama factory and the WHO’s missing health targets
- A multi-topic tech news roundup covering AI industry, energy, legal, and ethics stories - not a focused item.
- Includes notes on US-China AI safety talks and Anthropic's $30B funding round, but lacks depth on either.
- Low signal for APS readers; no Australia-specific content or actionable governance material.
-
The Download: deepfake porn’s stolen bodies and AI sharing private numbers
- A multi-topic tech news roundup covering AI, geopolitics, energy, and industry gossip in brief snippets.
- No single item develops enough depth to inform APS AI governance, strategy, or policy work.
- Low signal for APS readers; included for completeness rather than priority.