OpinionJuly 23, 2026 — 12:01pmAmid a raft of impressive numbers in Alphabet’s second-quarter results, one stands out. While quadrupling its profits, Google’s parent company reported the first quarter of negative free cash flow in its history.The apparent massive lift in profit overstated Alphabet’s operating performance by including a $US98 billion ($140 billion) non-cash mark-up in the value of its investments in SpaceX and Anthropic, but the cash-burn in the quarter was very real.The cash burn is real: Alphabet has spent up big to make Google an AI giant.GettyIt has become a common feature of big technology companies’ results as they plough ever-increasing amounts of cash and capital into artificial intelligence.In fact, both of those features of Alphabet’s numbers are common to the “hyperscalers” chasing their places in an AI-dominated future and share a common thread: the demands for capital to fund AI developments outstrip the ability of the companies to fund them internally.Increasingly, even the biggest tech companies are calling on debt markets, raising equity from the sharemarket and investing in their rivals to create daisy chains of mutual dependence – and vulnerability.The demands aren’t diminishing. Alphabet had capital expenditures of $US44.9 billion in the quarter – more than double the amount invested the same quarter of last year, and about 25 per cent more than it spent in the first quarter.The falls in Alphabet and Tesla’s share prices after their results suggest that investors’ patience and tolerance for the ever-escalating costs of their AI plans are starting to wane.It had previously forecast capex of $US180 billion to $US190 billion for the full year, but has now raised that to $US195 billion to $US205 billion.Alphabet wasn’t the only big tech company reporting results where AI investing overshadowed the performance of its more established operations.Elon Musk’s Tesla also released its June-quarter results this morning, which showed profits stayed essentially flat at $US1.1 billion despite a 25 per cent jump in revenue.Tesla generates more than 70 per cent of its revenue from sales of its electric cars – it sold a lot more of them in the quarter, albeit with significantly lower margins --but it is trying to transition to an AI-facilitated future of Robotaxis and robots.It, too, experienced negative cash flows in the quarter as it ploughed $US5.8 billion into spending, much of it AI- and robotics-related. That was more that twice its $US2.5 billion of capex in the first quarter, but well short of the rate needed to meet the $US25 billion of investment Musk now says he is targeting for the full-year.For both Alphabet and Tesla there are clear signs of surging AI-related sales growth.Alphabet’s cloud business revenues surged 82 per cent and its increasingly AI-integrated search unit also seems to be benefiting, while subscriptions to Tesla’s “Full-Self-Driving” automated driver-assist software, which costs $US99 a month, grew 15.6 per cent to 1.5 million. That’s a 56 per cent increase on their level a year ago.Both companies’ problem, and that of the entire AI sector, is that revenue growth rates are being outstripped by the growth rates in the spending required to keep up with the development of AI.Not only is the competition in AI increasingly being driven by the hardware – semiconductors, servers, data centres – but the intense competition and the demand it is generating is straining the capacity of the industry’s “picks and shovels” companies to deliver.The costs of “compute” are rising inexorably and sharply, and the timelines between committing the capital and generating revenue from it are starting to blow out.The falls in both companies’ share prices after their results (Alphabet was down 1.5 per cent and Tesla dropped 1.3 per cent) suggest that investors’ patience and tolerance for the ever-escalating costs of their AI plans are starting to wane.It’s not helping the companies or the environment in which they are seeking to raise capital that China seems to have a proliferation of open-source AI companies that can deliver near-comparable performance at a fraction of the cost of the hyperscalers, although the established tech giants do have the advantage of being able to integrate AI into their existing businesses and customer bases.Alphabet, like the other US hyperscalers (Amazon, Microsoft, Meta, Oracle), used to be a “capital-light” business.Today it and its peers are increasingly capital-intensive as the impact of the $US750 billion-plus they plan to spend this year – and the $US1 trillion-plus next year – continues to hit their balance sheets. Their financial profiles are being transformed, and their balance sheets are becoming leveraged.For instance, Alphabet’s total asset base has grown from $US595.3 billion a year ago to $US922 billion, with its long-term debt more than doubling from $US46.6 billion to $US98.2 billion to help fund that expansion. It also raised $US50 billion selling new shares as part of a planned $US80 billion capital raising.At least the hyperscalers, and Tesla for that matter, have cashflows and earnings from their pre-existing businesses to help fund their AI investments.Start-ups like OpenAI and Anthropic don’t have those legacy cashflows and therefore are reliant on regular equity injections from investors (including their competitors) and, increasingly, bond issues.The sharemarket performance of Elon Musk’s SpaceX sounds a warning suggesting that not all things AI always rise in value.Getty ImagesThere’s a circularity to the relationships between the industry participants that has so far enabled valuations to keep rising, which provides the necessary access to equity and debt markets (and paper profits from shareholdings in AI-driven companies).The sharemarket performance of another one of Musk’s companies, SpaceX, however, sounds a warning note suggesting that not all things AI always rise in value.SpaceX debuted on the sharemarket last month at an initial public offering price of $US135 a share. The stock opened above $US150, spiked above $US200 within its first fortnight of trading – and is currently priced at about $US115.The company has lost more than $US1 trillion of its peak market valuation and, with nearly another billion of “insider” shares coming out of lockup early next month, is at risk of falling further.SpaceX’s early market performance was an illustration of investors’ willingness to capitalise visions of what AI might be.Its subsequent fall signals that they also want some evidence that the vast amounts being invested in AI today will produce a return, commensurate with the risks, at some point in the not-too-distant future.The Business Briefing newsletter delivers major stories, exclusive coverage and expert opinion. Sign up to get it every weekday morning.More:AIOpinionGoogleTeslaElon MuskOpenAIAnthropicFor subscribersSpaceXEarnings seasonFrom our partners