Anthropic reveals revenue jump before listing and tests the resilience of infrastructure spending

At the moment when semiconductor stocks fell under the weight of doubts surrounding the viability of investment in data centers, Anthropic, the AI startup, chose to put its financial numbers on the table of investors in San Francisco. The company, developer of the "Claude" model, disclosed in a closed briefing on Wednesday a 45 % jump in revenue to $16.8 billion for the three months ended September 30, a step that paves the way for listing its shares on the stock exchange before the end of the current year.

The revenue trajectory reflects an accelerating quarterly growth that started at $4.7 billion in the first quarter, rose to $11.6 billion in the second quarter, and reached the third-quarter total.The rise was not limited to sales, as the company told investors at its headquarters that it had posted operating profit for the second quarter in a row, while declining to comment officially through its spokesperson. Those figures, together with data from its rival OpenAI, have become a key metric by which the market gauges generative-model makers’ ability to cover the massive computing costs that consume infrastructure budgets.

The financial disclosure coincided with a sharp sell-off that hit global technology markets on Wednesday, after remarks by Hock Tan, chief executive of Broadcom in Singapore, raised traders’ concerns by likening excessive investor expectations to gambling in casino halls. Shares of chip makers Nvidia, AMD and Intel fell on Wall Street, while Japan’s SoftBank slipped more than 9 % on Thursday morning, and Samsung dropped 9.4 % and SK Hynix 10.8 % in South Korea, prompting the KOSPI and KOSDAQ exchanges to suspend trading temporarily after declines exceeded eight %.

U.S. inflation data released on Wednesday added pressure, showing a September consumer-price increase that outpaced expectations, which raises the likelihood that the Federal Reserve will raise borrowing costs and adds to the financing burden of computing firms. At the same time, a marked timing divergence is evident between the two poles of generative-model makers; while Anthropic, the AI startup, is aiming to list this year despite its chief executive Dario Amodei’s calls to slow development because of risks, Sam Altman, chief executive of OpenAI, said in September that his company would not go public this year to address safety concerns, even though a future offering is planned that is expected to be valued at more than $1 trillion.

The shift of foundational models from a cash-burn phase to quarterly profitability changes the calculus of institutional adoption for technology decision-makers in the region.When a company like Anthropic, the AI startup, which expanded its international presence by opening an office in Singapore on Wednesday, proves its ability to generate billions of dollars from its operating model, the risk of relying on APIs diminishes for banks, telecom operators and government entities in the Gulf and Egypt; the landscape is no longer that of an emerging lab dependent on continuous funding rounds to pay its cloud bill. Yet the simultaneous downturn in chip supply chains and rising interest rates forces local infrastructure managers to scrutinize the cost of expanding servers and sovereign computing centers, because processor pricing will no longer be insulated from global capital costs.

We have observed in the trajectory of these figures that the AI battle is moving from showcasing theoretical capabilities to testing financial statements and the ability of commercial applications to offset billions of servers. Anthropic’s anticipated public offering will constitute the real test of public-market appetite, as profit margins and demand sustainability will be subjected to scrutiny that leaves companies with no refuge beyond actual accounting books.