AI use cases worth understanding.
Original explainers on documented AI use across Australian workplaces. The evidence timeline starts with research conducted through January 2025 and follows published cases into 2026. Each article shows its source date and what the source does, and does not, establish.
Use cases, with the details left in.
This library follows research and published cases from early 2025 through 2026. Each card shows when its source evidence appeared. Articles carry their actual publication date, not a backdated one.
People and work
How AI changes roles, participation and workplace experience.
Small business AI adoption starts with work design, not a software list
Jobs and Skills Australia's 2025 case compendium describes small firm barriers and workplace factors. Turn those observations into a practical adoption map.
Read the article People and workInclusive AI is a design decision, not a final accessibility check
Jobs and Skills Australia's 2025 case studies cover disability, Indigenous language and community co-design. Here is how to turn that into a respectful project brief.
Read the article People and workSEEK voice reference checks: faster completion still needs human care
SEEK reports that its voice AI reference checks cut completion time by 50 percent. Here is how to think about consent, identity, accessibility and review in hiring workflows.
Read the articleCustomer and operations
Use cases that improve customer responses and operational decisions.
Bunnings Team Member Assist: AI that starts with a better answer at the shelf
Wesfarmers describes an internal assistant giving Bunnings team members faster access to product and operational knowledge. The useful lesson is knowledge quality and escalation.
Read the article Customer and operationsCustomer service AI needs a handoff plan before it needs a personality
A practical design for routine enquiry automation, grounded in Wesfarmers' reported contact-centre use cases and the questions teams should answer first.
Read the articleHealth and care
Clinical and care workflows where review, consent and safety matter.
A 68 page care report becomes a three page shift brief
Regis Aged Care describes an AI assistant that summarises daily care reports for clinical managers. The important design question is how to surface urgent signals without hiding source detail.
Read the article Health and careSilverchain's home care AI roadmap starts with frontline problems
Silverchain's case describes voice documentation, early change detection and routine appointment communication. The key lesson is to label deployed and exploratory work separately.
Read the articleProfessional work
Document heavy work where AI can help people find, organise and review evidence.
Risk and compliance
Traceable workflows for risk, controls and regulatory change.
Ashurst's risk review: document extraction is useful when experts get the flags
Ashurst reports a 58 percent faster review for a major bank's collateral and loan documents. See the workflow, the claim's source and the controls that make it reviewable.
Read the article Risk and complianceCyber assessments and regulatory change: make every AI finding traceable
PwC's case stories describe AI mapping evidence to cyber controls and turning regulatory change into actions. The common requirement is traceability back to source.
Read the articleOperations and safety
AI in complex physical systems, with people responsible for the response.
Stanwell's energy models: why scenario speed matters more than a single forecast
Stanwell's reported AI platform models operational, weather and market information across thousands of scenarios. The case shows what decision support can look like in a complex energy system.
Read the article Operations and safetyBHP's object detection: move the hazard decision upstream
BHP's reported conveyor vision system looks for oversized rocks and foreign material before crushers and port systems. Learn how to frame a safety use case around prevention and escalation.
Read the articlePrivate AI
How to choose and govern the workspace where AI meets business information.
Public sector
Field notes and practical use cases from Australian organisations.
Policy work and AI: start with the question, not the prompt
What early Australian research with senior public servants reveals about using generative AI in policy work, and how to test it without outsourcing judgement.
Read the article Public sectorThe Treasury Copilot trial: useful work was mostly ordinary work
Australia's Treasury trial found promise in basic administrative tasks, alongside limits for complex work. Here is how to turn that into a sensible pilot.
Read the article Public sectorOne third used Copilot daily. What the APS trial actually tells us
The whole of government Copilot evaluation reported common tasks, perceived time savings and real barriers. Read the measures before borrowing the conclusion.
Read the articleRisk and safety
Field notes and practical use cases from Australian organisations.