The VET Futures Atlas
Live β€” TDA 2026
πŸ‡³πŸ‡Ώ New Zealand
System scale
~170 PTEs + 16 ITPs
Reform era
RoVE β€” ongoing
Qualification body
NZQA / Workforce Development Councils
AI adoption signal
Early / practitioner-led
Completion pressure
Moderate, rural access gap
Human Over
The Loop
The useful AI question in VET is not chatbots or cheating. It is who controls the qualification machinery β€” and whether AI makes that machinery better.
Power of TAFE → Brisbane → May 2026
πŸ‡¦πŸ‡Ί Australia
System scale
59 TAFEs + large PTE sector
Reform era
Jobs & Skills Councils β€” active
Qualification body
ASQA / JSCs by sector
AI adoption signal
Patchy / growing fast
Completion pressure
High β€” political priority
Convention
TAFE Directors Australia 2026
Theme
Power of TAFE
Dates
5–7 May 2026
Venue
Brisbane Convention & Exhibition Centre
Session 1A
Wed 6 May • 2:15–3:30pm
Plaza Ballroom
AI-generated assessments: enhancing personalisation and quality in TAFE delivery
Karl Hartley (Epic Learning / hotloop.ai) & Eve Price (Solutionary)
Field signals
What I'm watching at TDA
Five questions I'll be listening for across sessions, corridors, and conversations. These are the indicators that matter beyond the headline themes.
01
Governance
Who owns the qualification decision when AI drafted it?
Most AI assessment conversation skips liability entirely. I want to hear who is holding that.
02
Quality
Is moderation getting faster or just getting cheaper?
Speed and cost savings are easy to pitch. Evidence of maintained or improved judgement quality is rarer.
03
System design
Are TAFEs redesigning assessment architecture, or patching the existing one with AI?
The patch approach fails at scale. Architecture-first thinking is the tell.
04
Evidence
What does a strong evidence portfolio look like when AI helped build it?
Industry referees and authentic contexts still matter. Is anyone building frameworks for this?
05
Workforce
Are trainers being prepared for authoring roles, not just delivery roles?
If AI shifts work upstream to design and QA, the trainer professional development pipeline needs to shift with it.
Live journal
Field notes
Updated across the three days. Brief, sharp, real. Return here during and after the convention.
2026-05-04 — Pre-conference
Pre-arrival
Arrived with one clear question: does the sector's AI conversation happen at the chatbot layer or at the qualification machinery layer? Three days will tell.
2026-05-05 — Day 1
Watching
Opening plenary framing. Listening for: is AI framed as an efficiency tool, a quality tool, or something that changes what TAFE does structurally?
2026-05-06 — Day 2, 2:15pm
Session day
Today is the session. Eve Price and I in Plaza Ballroom, Stream 1A. The argument: AI hits the machinery first β€” standards, assessments, moderation, evidence, QA β€” and expert humans stay in control of each gate.
2026-05-07 — Day 3
Synthesis
Final day. Pulling together patterns from across sessions. What did the sector actually signal about readiness, resistance, and where the real reform energy sits?
Notes update during the convention. Refresh for latest signals.
Cross-Tasman lens
NZ vs Australia: where the systems differ
Same reform pressures. Different scale, velocity, and friction points. Useful to TAFEs thinking about what lessons transfer and what does not.
πŸ‡³πŸ‡Ώ New Zealand Dimension Australia πŸ‡¦πŸ‡Ί
~186 registered providers System scale 4,000+ registered providers
RoVE reshaping workforce dev councils Reform speed Jobs & Skills Councils accelerating
Strong micro-credential uptake Credential innovation Incremental, framework-cautious
Rural access, Māori equity gap Completion friction Socioeconomic, First Nations, remote
Practitioner-led, policy lagging AI readiness Sector patchy, TAFE brand advantage
NZQA / WDCs hold standard AI governance ASQA / JSCs β€” frameworks emerging
Learnable β€” small, fast to pilot Transfer value Scale to replicate NZ pilots nationally
Core argument
AI hits the machinery first
Before AI reaches learners, it reaches the infrastructure of qualification. Five stages β€” and who holds responsibility at each gate.
01
Standards Design
Human: intent & values
AI: gap analysis, drafts
›
02
Assessment Authoring
Human: validity sign-off
AI: generate, vary, align
›
03
Moderation Support
Human: final judgement
AI: consistency checks
›
04
Evidence Architecture
Human: context design
AI: portfolio scaffolding
›
05
QA & Compliance
Human: accountable sign-off
AI: audit trail, flags
Coral = human decisions & accountability
Teal = AI processing & support
vetinsights.hotloop.ai
For TDA attendees
Scan for live field notes & post-conference synthesis
  • Follow Karl's live signals from across the three days β€” updated in real time
  • Get the post-conference synthesis: what the sector actually signalled, and what it means for NZ and AUS
  • See the practical AI argument beyond the hype: qualification machinery, not chatbots
  • NZ vs AUS comparison built for practitioners, not policy documents
vetinsights.hotloop.ai
Post-conference
What this becomes after May 7
A field journal becomes durable intelligence. Return here in late May for the full synthesis β€” what the sector signalled, what it means, and what to do next.
πŸ“‘
Signal synthesis
Cross-session patterns distilled into 5–7 clear signals about where the VET sector's AI conversation actually is β€” versus where the presentations said it was.
βš–οΈ
NZ & AUS comparison update
Updated with conference evidence. Where do the two systems differ in AI readiness, governance appetite, and practitioner confidence?
πŸ”§
Practical argument refined
The machinery-first case, sharpened by what worked and what was challenged in the room. A resource practitioners can use β€” not just read.
πŸ—ΊοΈ
Where to go next
Concrete next steps for TAFE leaders, assessment designers, and VET policy people who want to act on what TDA surfaced β€” not just file it.