AI is everywhere in finance. Governing it is not.
A new report from the Actuaries Institute and UTS put numbers on the AI governance gap in Australian finance, and singled out credit decisioning as the place funds should start.
Hi folks,
This week brought fresh survey evidence of how deeply AI is already embedded across financial services.
The one thing.
AI is everywhere in finance. Governing it is not.
Published on 1 September, AI Risk Management in the Financial Services Sector walks through:
- Who is accountable for the AI risk
- How to classify it
- How to measure it and what controls fit
The survey behind it is the striking part:
- 93% of financial firms already use AI
- Fewer than half have risk-assessed their own use of it
- Fewer than one in three have a single AI item on their risk register
- Yet 74% still rate their AI risk management as effective
The report singled out loan-related workflows as a particularly high-risk case: get it wrong and you reject applicants unfairly, or discriminate.
What it means for a fund using AI:
You need to develop proper governance, especially around credit decisioning. AI risk should sit on a risk register, with a clear audit trail of what the model did and did not do, and where a human was in the loop.
In the mix.
- Anthropic released a new flagship model, Claude Fable 5.1 (Anthropic)
- Anthropic positions it as its most advanced model yet for coding and knowledge work, ahead of Claude Opus 5 on its own benchmarks.
- The signal is the capability jump, not the price. On agentic coding, the multi-step problem-solving that stands in for a lot of knowledge work, it went from 42% to 56% on one industry benchmark in a single release.
- Why it matters: more of the knowledge work your teams do is becoming feasible for AI to do, and the frontier is still moving in months, not years.
- Capital kept committing to private credit anyway (Investment Magazine)
- Four large Australian institutions committed over A$700 million to European middle-market direct lending through Nuveen's Arcmont.
- Why it matters: a week after ASIC's "first significant cracks", allocators are still writing direct-lending cheques. They are just asking harder questions about liquidity and evidence first, which is what the fortnight has been about.
Levercon builds the AI operating system for credit funds: we connect your data, deploy agents across origination and monitoring, and run the repetitive work end to end.
Book a strategy call.
See where AI fits in your firm.
A short conversation about where AI could help your firm, what to build first, and what it would take.