APRA’s late-April letter on AI was addressed to Banks, Insurers and Super Trustees. It wasn’t addressed to AFSLs – and Pete Worn reckons that distinction will matter for about six months.
In this episode, Danni and Pete unpack Pete’s latest Code & Capital on what APRA’s AI letter signals for the broader advice industry, and what ASIC is almost certainly preparing in the background.
They cover:
- Why AI isn’t “just another technology” – and why deterministic thinking gets you into trouble fast
- The four areas APRA put on notice: cyber, governance, supplier risk, change management
- Why boards are flying blind on AI literacy, and why the vendor lunch-and-learn is now a regulatory red flag
- The fourth layer of due diligence no one was doing before – and why it changes everything
- Fresh data from Finura’s latest Tech Health Check: only 57% of licensee-aligned practices run due diligence on tech providers, and 48.6% of practices admit they can’t explain every app in their own stack
- Claude Code in production, frontier AI vendors that can’t answer DD questions, and the systemic risks no one is mapping
- Australia’s surprisingly low AI trust scores, UHNW clients refusing meeting recordings, and what Scott Galloway calls “the biggest brand problem of any technology in human history”
A practical, sometimes uncomfortable look at the gap between how fast advice firms are adopting AI and how slowly governance is catching up.