From Singapore: When the chief executive of the world's most profitable bank invokes the global financial crisis as a cautionary tale, markets pay attention. JPMorgan Chase chief executive Jamie Dimon has issued a pointed warning about the current frenzy surrounding artificial intelligence, suggesting that reckless investment behaviour in the sector could trigger a wave of corporate loan defaults with echoes of 2008, as reported by the Sydney Morning Herald.
The comparison is not idle. Dimon expressed concern that the AI boom has encouraged a level of financial exuberance that looks uncomfortably familiar. "People doing dumb things" was his blunt assessment of some of the activity he is observing in capital markets tied to artificial intelligence.
His specific worry centres on software companies. The theory runs like this: if AI dramatically reduces the cost and labour required to build software, many existing software firms could see their revenue models collapse faster than their balance sheets can absorb. Companies that borrowed heavily on the strength of their current valuations may find themselves unable to service debt as their business models erode beneath them. Banks holding that debt then face the kind of credit quality deterioration that, in aggregate, can tip into systemic risk.
What It Means for Australian Investors
For Australian investors, the signal is harder to dismiss than it might initially seem. Superannuation funds have deployed significant capital into US technology equities over the past decade, chasing the sector's outsized returns. The Australian Prudential Regulation Authority supervises a superannuation system holding more than three trillion dollars in assets, a substantial portion of which carries indirect exposure to American technology stocks. A disorderly repricing of AI-adjacent software companies in US credit and equity markets would not stay contained to Wall Street.
Across the region, the trend is unmistakable. Asian technology giants, from South Korea's semiconductor producers to India's IT services firms, have tied portions of their growth narratives to the AI investment cycle. ASEAN's technology sector, still maturing but increasingly integrated into global supply chains, is watching developments in American AI financing closely. A credit event in US software lending would ripple through Asian venture capital and private credit markets, which have themselves expanded rapidly in recent years.
The Case for Optimism
There is, of course, a serious case to be made on the other side. AI optimists, including a number of credible economists and technology researchers, argue that Dimon's GFC comparison fundamentally misunderstands the nature of the current cycle. Unlike subprime mortgage lending, where the underlying asset consisted of inflated house prices with little genuine foundation, AI genuinely improves productivity across measurable dimensions. The technology is real, even if some valuations attached to it are not. Pointing to documented gains in software development speed, drug discovery timelines, and logistics efficiency, proponents argue the economy is in the early stages of a genuine structural shift, not a speculative bubble.
The Australian government has been broadly supportive of AI adoption while the Australian Securities and Investments Commission monitors the use of AI in financial services for consumer protection risks. The Reserve Bank of Australia has flagged productivity gains from technology as a meaningful variable in its long-run growth modelling, reflecting cautious institutional optimism rather than alarm.
Where does the weight of evidence fall? Probably somewhere between the two poles. Dimon is not predicting catastrophe; he is identifying a specific mechanism, over-leveraged software companies vulnerable to AI-driven disruption, as a credible pressure point in the credit system. That is a narrower and more defensible claim than a blanket assertion that AI is a bubble. His track record of reading credit risk is strong, even if his timing has occasionally been imprecise.
For Australian businesses and investors with exposure to global technology lending, the prudent response is not panic but preparation: reviewing counterparty credit quality, stress-testing portfolio assumptions, and resisting the pressure to chase late-cycle AI returns simply because peers have done so. The GFC parallel Dimon draws is instructive not because history is poised to repeat itself, but because the mechanisms of credit stress are remarkably consistent across cycles, whatever the underlying asset class happens to be.