AI implementation
Who owns an AI workflow after launch? A register for Australian businesses
A recommended register of who runs, watches, changes, pauses and exits an AI workflow after...
Research
Evidence for safer AI adoption.
Research and working notes on AI adoption, governance, privacy and safer rollout for Australian businesses.
The live index
Research for practical operating decisions.
14 published articles and working notes.
Editorial system
Five lanes for understanding agent work.
- GovernanceControl boundariesWhere agent work starts, stops, and stays accountable.
- SovereigntyAustralian data contextWhat local operating conditions mean for data and adoption.
- ApprovalEvidence and decisionsHow review points and retained evidence shape safer action.
- SecurityAgent exposureThe new boundaries that arrive when software can take action.
- AdoptionOperating systemsHow teams move from experiments to repeatable work.
The index
14 published1234
AI implementation
AI pilot acceptance checklist: proceed, revise or stop
A recommended test schedule for a document-to-draft AI pilot in an Australian business: approval, exception...
AI implementation
AI implementation costs in Australia: a buyer's worksheet
Compare AI implementation costs on the same scope with a blank lifecycle worksheet covering cash,...
AI implementation
How to choose an AI implementation partner in Australia
Compare AI implementation partners in Australia with supplier interview questions, acceptance tests, privacy checks and...
Cybersecurity
Claude Mythos and cybersecurity: turn the announcement into a patching decision
What Anthropic reported about Mythos Preview and Project Glasswing, and a risk-based checklist for the...
Enterprise AI
Claude Mythos Preview: procurement questions for an enterprise AI strategy
Separate a restricted model announcement from a deployable business service. Ask about access, data flow,...
Security & Compliance
Mythos and zero-day reports: an action checklist for software owners
Use the primary announcement and follow-up to ask better questions about affected software, vendor fixes...
Business Strategy
Mythos Preview: separate model capability from an enterprise commitment
A short decision note on access and procurement after Anthropic's Project Glasswing announcement, without speculative...
Industry Insights
Australian SME AI adoption: what the October–December 2024 data can tell you
The cited source covers Q4 2024, not Q1 2026. Separate reported adoption from measured return,...
Regulation
Australia's AI Regulation Gap: What the Lack of Mandatory Guardrails Means for Your Business
Australia has scrapped its AI advisory body, replaced its voluntary safety standard with guidance, and...
AI Trends
Sora's discontinuation: dates, exports and a continuity checklist
OpenAI distinguishes the ended web and app experiences from the scheduled API shutdown. Prepare exports...
Industry News
SoftBank's USD 40 billion bridge facility: what it does and does not mean for an AI buyer
The March 2026 disclosure concerned a financing facility, not a USD 40 billion OpenAI equity...
AI Agents
An AI agent can still look like a chatbot: inspect what it can do
The interface does not tell you the permissions. A short guide to separating conversation, tool...
AI Fundamentals
AI agents for Australian knowledge workers: what to delegate and what to keep under review
Choose between native AI features, fixed workflows and agents by the work, permissions, approval points...
Ready to map one real workflow?
Explore our capabilities