Know what good looks like
The four things APRA’s letter expects you to evidence, translated from supervisory language into a concrete checklist.
APRA letter to industry · 30 April 2026
APRA has set its expectations for AI.
Can you evidence them?
On 30 April 2026 APRA told regulated entities what good looks like for AI: a live inventory, behaviour monitoring, clear lifecycle ownership, and assurance a supervisor can rely on. We wrote the checklist of what you’ll actually be asked to show, and the gaps most firms still have.
The problem
A policy on paper is not evidence.
Most banks, insurers and super funds have an AI policy. Few can produce, on request, a current inventory of every model and agent in use, evidence that their behaviour is monitored, a named owner for each through its lifecycle, and assurance over the AI-driven decisions that affect customers. That gap between policy and evidence is exactly what supervision now probes.
What it does
The work, done for you.
Find your evidence gaps
Where most regulated firms fall short today: stale inventories, unmonitored agents, and assurance that stops at the policy.
A board-ready position
A clear read your board and CRO can stand behind, and act on, before the supervision questions arrive.
Free guide
Get the plain-English guide.
Free guide
APRA’s AI Expectations: 2026 Readiness Checklist
What APRA’s 30 April 2026 letter expects regulated entities to evidence for AI, the common gaps, and where to start closing them.
- The four expectations, in plain English
- The inventory, monitoring and ownership evidence to hold
- What assurance over AI decisions actually means
- The gaps to close before the next supervision cycle