The Reserve Bank of Australia warned on 1 October in its semi-annual Financial Stability Review that advancing AI could make cyberattacks on financial institutions easier, while growing debt-funded investment could expose lenders and investors to losses, News.Az reported.
Key points
- The RBA warned that frontier AI advances are reducing the cost and expertise needed for sophisticated cyberattacks.
- It said AI companies’ increasing use of debt is widening investors’ exposure in public and private markets.
- The bank warned that financing through special purpose vehicles could conceal exposures and make links between financial entities less transparent.
RBA warns of cheaper cyberattacks
The RBA said advances in frontier AI capabilities, alongside broader technological progress, are reducing both the expense and the technical knowledge required to mount sophisticated attacks. Its warning covers financial institutions and the infrastructure on which financial markets operate.
The central bank said AI could eventually improve productivity and operational resilience. It also warned that the pace of development could make the cyber threat environment more difficult initially and increase the risk that existing weaknesses in financial organisations’ operations are exposed.
Those are separate possible effects on the same institutions. The RBA’s warning concerns a period in which potential improvements to their resilience coexist with a greater risk of attack against weaknesses they already have.
AI debt reaches public and private markets
On investment, the RBA warned that the scale and pace of financing for AI could create financial stability vulnerabilities if they continue to expand over the coming years. Companies involved in AI are increasingly using debt for large projects, it said, extending investors’ exposure to the sector across both public and private markets.
The bank expects the industry to become more important in public and private credit markets as it issues large volumes of debt. It likened that expected expansion to a trend already seen in equity markets, with a wider range of investors exposed to AI investment.
The RBA said that, if AI project outlays rise in line with current market expectations, substantial external financing would increase exposure for banks, debt markets, private lenders and other investing organisations. The scale of that exposure therefore depends, in the bank’s account, on how much of the expected spending takes place.
The RBA governor has also said AI investment is adding to inflation before productivity, as AI Affairs previously reported. The Financial Stability Review addresses a different consequence of investment: who holds the credit exposure as projects are financed.
Special purpose vehicles could conceal AI exposures
Some major AI infrastructure projects are being funded through off-balance-sheet arrangements involving special purpose vehicles. The RBA warned that their growing use could leave exposures hidden and make connections between financial entities less transparent, News.Az reported.
The bank’s concern extends beyond the companies undertaking the projects. Banks, bond investors, private credit providers and other institutions could hold exposure as AI capital spending expands, while special purpose vehicles could make some of the connections between those financial entities harder to see.
The returns expected from that spending rest on forecasts of widespread AI adoption and strong revenue growth, the RBA said. It cautioned that the timing, size and distribution of those gains remain uncertain. Borrowing for large projects can therefore expand while the earnings assumed in long-term forecasts remain dependent on future adoption.
The central bank said returns below expectations could potentially cause losses for lenders and investors. That warning applies to financing whose reach is expected to grow in credit markets, including the off-balance-sheet arrangements through which the RBA said exposures could be hidden.
