Elon Musk has lost his attempt to neuter OpenAI. That’s probably a good thing for the entire artificial intelligence complex, including Musk’s own SpaceX.
On Monday, a jury in the trial Musk kicked off against OpenAI in 2024 found that Musk had launched the legal action after the timeframe allowed by the law. The judge in the case concurred with that finding, which Musk said he would appeal.
Musk had alleged that OpenAI co-founder Sam Altman and president Greg Brockman “stole a charity” when they restructured the OpenAI into a for-profit entity last year. Musk was seeking up to $US150 billion ($210 billion) in damages, called for the reversion of OpenAI to a not-for-profit and wanted Altman and Brockman ejected from the company.
Had he succeeded, it would have gutted OpenAI, cutting it off from the massive flows of funding required to train its models and acquire the chips and access to data centres that would enable it to remain in the frenetic scramble to develop ever-more sophisticated AI tools.
It would also have had significant, and threatening, implications for the rest of the sector, with OpenAI at the centre of the complex agreements and funding deals that the AI ecosystem has developed to raise the ever-increasing amounts of capital required to remain competitive.
If OpenAI had lost the case and been removed from its pursuit of artificial general intelligence, or human-levels of artificial intelligence, it would have sent shockwaves – and massive losses – throughout the sector.
If OpenAI had lost the case and been removed from its pursuit of artificial general intelligence, it would have sent shockwaves – and massive losses – throughout the sector.
It would also have derailed OpenAI’s future funding capacity. OpenAI, which raised $US122 billion at a valuation of $US852 billion in March, has tentative plans for an initial public offering (IPO) later this year to raise another $US60 billion or more, at a share price that would value the business at $US1 trillion-plus later this year.
Coincidentally or otherwise, Musk’s SpaceX will this week issue a public prospectus for a capital raising of its own, seeking to sell investors $US75 billion to $US80 billion worth of shares in an IPO that would value the company at $US1.75 trillion to $US2 trillion. That offering would have benefited had OpenAI been taken out of the race to develop AI and been withdrawn from the competition for funding.
In what shapes as a critical year for AI, given its insatiable appetite for capital, Anthropic is also mulling an IPO of its own.
The company, at the leading edge of AI development, is reportedly seeking to raise $US30 billion from investors at a $US900 billion valuation now, with an IPO earmarked for later in the year, presumably at an even higher valuation. It raised $US15 billion from Google and Amazon in February, with performance-conditional commitments for a further $US50 billion, at a $US350 billion valuation.
The rate at which AI company valuations are escalating reflects the pace of the industry’s development, and the near mania surrounding everything AI-related.
Anthropic, which had an annualised revenue run-rate of $US9 billion at the end of last year, recently said it was now generating revenue at an annualised rate of $US30 billion. That illustrates how rapidly the commercialisation of AI is occurring, albeit that Anthropic – like SpaceX and OpenAI – is still unprofitable.
The potential IPOs of the three big players this year, with combined valuations ranging between $US3.5 trillion and $US4 trillion, will be a massive test of the sharemarket’s capacity and faith in the sector’s ability to eventually produce the mega returns on capital required to justify those valuations and the hype around the sector.
At present, AI companies are devouring capital at a rate that’s even faster than their revenue growth. AI might present as software, but it is proving to be an extremely capital-intensive sector.
The fate of the companies’ fundraising is critical not just for AI development and the companies doing that development, but for the wider sharemarket.
The half dozen or so big tech US companies that are investing the massive sums required – an estimated $US700 billion-plus this year, and more than $US1 trillion next year – represent more than one-third of the entire US sharemarket’s value.
They’ve been the driving force behind the market’s 87 per cent gain since the launch of OpenAI’s ChatGPT in November 2022 ignited the frenzy around AI.
Should any of the planned IPOs fail to deliver on expectations, it would send a shudder not just through the sector, but the entire market, which has become increasingly vulnerable to even the slightest AI setback.
The market is also vulnerable to the external economic environment, where Donald Trump’s tariffs and the war on Iran have caused inflation rates in the US and elsewhere to spike and growth rates to slow.
The inflationary pressures from the oil shock in particular will continue to rise as the higher prices for gasoline and diesel work their way deeper into economies, product costs and eventually prices.
Higher interest rates – and the cost of traded debt is already rising – could puncture the booming sharemarket.
With AI companies increasingly testing the limits of the equity market’s capacity to supply capital on the scale required, they have resorted to debt markets to fund their insatiable needs for new capital.
Higher interest costs, or a reduced willingness to lend, would not just hurt the economics of AI, but could constrain development in a period where the sector is dependent on continuing rapid progress to justify its valuations and maintain its access to capital.
There’s a set of dependencies in the sector which, if they were to reverse, would be extremely destructive beyond the AI industry.
While there are lurking threats to the perpetual mega fund-raising that the AI companies require to maintain momentum, there’s also continuing investor enthusiasm and optimism about their future.
Last week, a relatively small AI chipmaker, Cerebras Systems, listed in the US. Its IPO price was $US185 a share, after it lifted it from an initial $115 a share. The stock opened at $US350 and closed at $US311 a share and was still trading at just under $US300 a share on Monday.
That tends to underscore that the frenzy around AI stocks and everything related – Cerebras makes the advanced computer chips used to train AI models – is, for the moment at least, continuing unabated and unaffected by the economic and geopolitical risks Trump has created.
For SpaceX, OpenAI and Anthropic, as they pursue their IPOs this year, the defining questionmark will be whether the hype and investor enthusiasm for all things AI are sustained until they can get the more permanent and continuing access to capital in place that a sharemarket listing provides.
Musk, who co-founded OpenAI and left the company after a power struggle with Altman and Brockman has, for the moment at least, lost the opportunity for revenge and for damaging a SpaceX competitor.
For OpenAI, the AI ecosystem and the sharemarket at large, that’s at least one threat less.
The Market Recap newsletter is a wrap of the day’s trading. Get it each weekday afternoon.
Stephen Bartholomeusz is one of Australia’s most respected business journalists. He was most recently co-founder and associate editor of the Business Spectator website and an associate editor and senior columnist at The Australian.Connect via email.
