Anthropic’s net loss widened sharply to $80.7 billion in the first half of 2026, driven largely by an accounting charge linked to convertible bonds as the artificial intelligence startup prepares for a potential initial public offering (IPO) in November. The disclosure has raised fresh questions about the company’s valuation and the enormous costs associated with developing and deploying advanced AI systems.
According to a draft IPO prospectus reviewed by Reuters, the San Francisco-based company recorded a loss of $80.7 billion in the January-June period, compared with $13 billion in the corresponding period a year earlier. However, much of the increase did not stem from cash expenses incurred through its day-to-day operations.
Instead, approximately $72 billion of the reported loss came from an accounting charge reflecting an increase in the estimated fair value of bonds that could eventually convert into Anthropic shares.
The disclosure comes as investors reassess expectations surrounding the company’s public listing, with some reportedly considering a valuation of around $1.5 trillion following concerns about revenue growth and mounting expenditure.
Credits: Reuters
$72 Billion Accounting Charge Drives Anthropic’s Loss
The bulk of Anthropic’s first-half loss was attributable to changes in the estimated value of convertible bonds, rather than direct operational spending. Such instruments can become more expensive on a company’s financial statements as its valuation rises, depending on the terms of the securities and the applicable accounting treatment.
The $72 billion charge therefore does not represent an equivalent cash outflow during the first half. Nevertheless, it significantly affected the company’s reported financial performance and highlights the complexities involved in evaluating fast-growing AI companies with substantial equity-linked financing arrangements.
Anthropic had also reported a net loss of $42 billion in 2025, with much of that figure attributed to a similar accounting charge, according to a Reuters report published in September.
Excluding the accounting charge, Anthropic recorded an operating loss of approximately $8.1 billion on revenue of $4.6 billion in 2025, underscoring the substantial costs involved in building and commercialising its AI technology.
Anthropic’s Computing Spending Surges
Beyond accounting adjustments, Anthropic’s infrastructure requirements remain a significant financial consideration. The company spent nearly $9 billion on computing capacity and infrastructure in the first half of 2026, compared with $2.3 billion in the same period last year.
The spending reflects the resources required to train, operate and improve increasingly capable AI models, which depend on large quantities of computing power, specialised chips and data-centre capacity.
Anthropic has committed to spending at least $518 billion over a decade on cloud services, computing and infrastructure. These commitments highlight the scale of investment required to compete in a market where technology companies are racing to develop more powerful models and expand access to AI tools.
The challenge for Anthropic and its competitors is to ensure that growing demand for AI products translates into sustainable revenue and eventually offsets the cost of developing and operating these systems.
Revenue Growth Raises Questions Ahead of IPO
Anthropic’s financial disclosures present a mixed picture for prospective investors. At a meeting with IPO investors on October 14, the company reportedly said its third-quarter revenue rose 45% sequentially to $16.8 billion, according to Reuters, which cited people familiar with the matter.
The increase followed revenue of $11.6 billion in the second quarter, when sales had more than doubled compared with the first quarter.
Although the latest figures demonstrate substantial commercial momentum, some investors were reportedly disappointed by the slowdown in sequential growth. This has contributed to questions over whether the company can justify its anticipated public-market valuation while continuing to spend heavily on infrastructure.
Anthropic’s planned November IPO is also expected to draw attention across the broader AI industry. A weaker reception could influence investor sentiment towards other AI companies, including OpenAI, which could pursue a public listing next year.
Claude Code Emerges as a Growth Driver
One of Anthropic’s important commercial products is Claude Code, its AI-powered coding assistant that helps developers write software, build applications and automate programming tasks.
The product has benefited from growing interest in AI tools that can perform practical work rather than simply generate text or answer questions. Coding assistants are increasingly being integrated into software development workflows, creating opportunities for AI companies to generate revenue from enterprise customers and individual developers.
However, continued adoption will need to translate into durable revenue growth to support Anthropic’s long-term investment commitments.
Credits: Reuters
What Anthropic’s IPO Means for the AI Market
Anthropic’s planned public listing comes at a crucial moment for the AI sector, as investors increasingly scrutinise the relationship between technological progress, revenue growth and infrastructure spending.
The $80.7 billion first-half loss may appear alarming at first glance, but the $72 billion accounting charge means the headline figure requires careful interpretation. At the same time, the company’s infrastructure expenditure, operating losses and long-term commitments remain important considerations for investors.
As Anthropic moves towards its anticipated IPO, prospective shareholders will be watching revenue growth, operating costs and the commercial performance of products such as Claude Code. The listing could become an important test of whether public-market investors are willing to support the enormous capital requirements of the AI race.
