Anthropic said it has filed confidential paperwork with the US Securities and Exchange Commission to hold an initial public offering, joining a wave of highly anticipated flotations expected to reshape Wall Street. The AI company, known for its Claude family of large language models, revealed the news amid what could be an unprecedented year for technology listings.
SpaceX, which holds a virtual monopoly on US space launches, has also filed for what is widely expected to be a record-breaking IPO. Meanwhile, OpenAI, one of Anthropic's closest rivals, is reportedly planning its own offering later in 2026. The simultaneous moves signal a new phase in the AI arms race, as private companies seek deeper access to public capital markets to fund expensive infrastructure and research.
Confidential filing process
The confidential IPO process allows companies to submit draft registration documents to the SEC without immediately exposing sensitive financial details to the public. Companies use this approach to gauge regulatory feedback and market conditions before committing to a formal public debut. Once a company begins an active roadshow and releases its prospectus, its financial information becomes available to investors and the public.
Anthropic's decision to file confidentially does not necessarily mean an immediate IPO. The timeline depends on SEC review, market conditions, and the company's readiness to disclose its operations to public investors. Still, the filing represents a major milestone for a company founded only in 2021 and now considered one of the most important players in generative AI.
Rapid growth since founding
Anthropic was established by former OpenAI executives and researchers, including Dario Amodei and Daniela Amodei. The founders left OpenAI in 2020 and launched Anthropic with an emphasis on AI safety, interpretability, and long-term alignment. The company chose a public benefit corporation structure, aiming to balance commercial interests with responsible development.
In just four years, Anthropic has grown from a research-focused startup into a major commercial force. Its Claude chatbot and associated developer tools are used by a broad range of businesses, from startups to large enterprises. The company's relative focus on tools for programmers and business users has set it apart from some competitors that have pursued consumer-facing products more aggressively.
Claude model momentum
Last autumn, the company's Claude Opus 4.5 model began to gain significant popularity, especially with software engineers. Developers praised its ability to handle long contexts, generate high-quality code, and integrate with existing workflows. The success of Claude Opus 4.5 helped drive Anthropic's revenue growth and expanded its presence in the competitive AI coding market.
More recently, Anthropic received widespread attention after claiming that its Claude Mythos model was significantly more capable than other tools at finding security flaws in computer systems. The company said it worked with banks, government agencies, and other security-sensitive organizations to help harden their systems before the model's public release. This emphasis on cybersecurity reinforced Anthropic's positioning as a trusted provider for high-stakes users.
Claude Mythos was described as a major leap in the AI model's ability to reason about complex codebases and identify vulnerabilities that traditional scanning tools might miss. Early tests reportedly showed the model outperforming existing automated security tools and, in some cases, matching results from human experts. The model's capabilities raised both excitement and concern, as security researchers considered how such an AI could be used by malicious actors. Anthropic said it took extensive measures to limit dangerous use of the technology, including safeguards, access controls, and partnerships with defensive security teams.
Revenue surge and first profit
The growing adoption of Claude models has translated into a sharp rise in revenue. Last month, multiple reports indicated that surging revenue may see Anthropic record its first profitable quarter during the period ended in June. According to these reports, income for that quarter was on track to more than double over the previous quarter. If realized, the profit would be a symbolic milestone for an AI industry known for heavy losses and massive capital burn.
But the company reportedly does not necessarily expect to show a profit in future quarters. Anthropic is also ramping up spending on compute infrastructure, including data center capacity, chips, and energy. The AI sector currently operates in an environment where success requires not only superior algorithms but also enormous computational resources. Model training and inference at scale consume vast amounts of electricity and require access to advanced semiconductors.
Anthropic's spending commitments highlight the structural challenge facing AI companies: rapid growth is often accompanied by even faster growth in costs. Public investors may need to recalibrate their expectations if Anthropic reports profits in some quarters and losses in others. The company's leadership has signaled that it will prioritize long-term capability gains and infrastructure investment over short-term profitability.
Massive data center deal
That approach is evident in a recent deal with SpaceX to use capacity at its Colossus 1 and Colossus 2 data centers. Anthropic is reportedly paying $1.25 billion (£930 million) a month for this capacity, an extraordinary sum that illustrates the scale of modern AI infrastructure. The agreement provides Anthropic with dedicated access to powerful computing resources and helps lock in capacity amid a global shortage of data center space.
The Colossus data centers are part of a broader trend in which AI companies seek strategic partnerships with infrastructure providers. By securing long-term capacity, Anthropic aims to avoid bottlenecks that could slow its training schedules. However, the monthly payment also places significant pressure on the company's balance sheet, reinforcing its need to generate high revenue growth.
The deal with SpaceX is notable not only because of the price but also because it brings together two private companies that are both preparing for IPOs. SpaceX has its own ambitions to list on public markets, and its data center business could make it a critical player in the AI ecosystem beyond its traditional role as a launch provider.
Valuation surpasses OpenAI
Anthropic's most recent funding round valued it at $965 billion, roughly $1 trillion shy of becoming one of the most valuable companies in the world. The valuation surpassed OpenAI for the first time, reflecting the market's growing confidence in Anthropic's commercial strategy and technical direction. OpenAI was most recently valued at around $852 billion after its own fundraising efforts.
The valuation gap may shift again as both companies pursue capital ahead of their expected listings. OpenAI has broad consumer recognition through ChatGPT, while Anthropic has built a reputation for safety-conscious enterprise AI. Each company has strengths in different areas, and their valuations will likely be tested in public markets where investors demand stronger evidence of sustainable profitability.
Industry challenges
Despite widespread enthusiasm around AI, industry analysts note that the technology has not yet progressed to the stage of being broadly useful for day-to-day tasks. Gartner has estimated that it could be two to five years before technologies such as foundational models or AI agents become truly productive and mainstream. The gap between current demonstration capabilities and reliable, cost-effective deployment remains a key concern for enterprise buyers.
Many organizations have experimented with AI assistants, but few have fully integrated them into daily workflows. Tasks that are routine for humans, such as handling complex customer service inquiries or managing multi-step business processes, remain difficult for AI systems to perform without error. Companies like Anthropic are investing in orchestration, evaluation, and guardrails to make AI agents more dependable, but widespread adoption may take years.
The AI industry also faces regulatory uncertainty. Governments in the United States, Europe, and Asia are considering rules that could affect how AI models are developed, deployed, and audited. Anthropic has positioned itself as a proponent of responsible AI and has supported some regulatory measures. Public ownership would bring greater scrutiny from investors, policymakers, and civil society groups.
A crowded IPO pipeline
The prospect of Anthropic and SpaceX going public is already generating excitement among investment banks and institutional investors. A successful Anthropic IPO would test whether the market's appetite for AI stocks extends to companies with future profitability uncertain. It would also provide a benchmark for other AI developers considering listings.
OpenAI is reportedly planning an offering later in 2026, which could create a rare competitive dynamic in which the two most important generative AI companies trade on public markets at roughly the same time. Investors would have the ability to compare their financial disclosures, growth metrics, and management strategies directly. That transparency could lead to more informed valuation decisions but also exposes both companies to intense quarter-by-quarter scrutiny.
Neither SpaceX nor OpenAI is profitable, underscoring how these listings are based on future expectations rather than current earnings. SpaceX has a dominant position in launch services and has generated substantial revenue, but its high costs and ambitious projects have kept it in the red. OpenAI, similarly, has spent heavily on training models and compute capacity as it pursues artificial general intelligence.
Anthropic's confidential filing is the latest signal that the company is preparing for one of the most closely watched IPOs in technology history. Its success will depend on sustaining its revenue growth, managing its enormous infrastructure costs, and convincing investors that AI models can become consistently profitable businesses. With data center deals, advanced models, and a valuation near $1 trillion, Anthropic is entering the public-market arena at a moment of both extraordinary opportunity and significant uncertainty.
Source: Silicon UK News