Anthropic files for massive IPO with 93-page AI risk disclosure

A massive market debut

Anthropic has officially filed for an initial public offering (IPO), marking a major milestone for the San Francisco-based AI lab. The company intends to list on the Nasdaq exchange under the ticker ANT, with a potential debut as early as November 16. This move could make it one of the largest stock market debuts in history, with a reported valuation of up to $2 trillion.

While the company has not yet disclosed the specific number of shares it plans to sell or the final price, the filing sets the stage for a high-profile entry into the public market. The move comes after Anthropic was last valued at $965 billion following a $65 billion funding round in May. This transition signals a shift toward facing heightened public and regulatory scrutiny as a public entity.

The filing serves as a litmus test for Wall Street's appetite for artificial intelligence. It arrives at a time when other major AI players have hesitated to go public due to concerns over debt-fueled infrastructure costs and uncertain returns. By moving forward, Anthropic is positioning itself as a leader in a race where global spending on AI services and infrastructure is expected to hit $3.6 trillion by 2027.

Surging growth and losses

The financial details in the filing reveal a company experiencing explosive growth alongside significant accounting complexities. Anthropic reported that its revenue grew 1,300% year over year during the first six months of 2026, reaching $16.3 billion. This growth is primarily driven by its Claude AI system, which has gained significant traction among enterprise customers.

Despite the revenue surge, the company's books showed a total deficit exceeding $80 billion. However, the filing clarifies that only a small fraction of this figure represents actual spending. More than $72 billion of that loss was attributed to changes in the fair value of convertible notes, derivative liabilities, and SAFEs, which are types of investment contracts.

The remaining portion of the loss was largely tied to infrastructure and computing costs, which totaled just under $9 billion. Anthropic also highlighted its internal efficiency, noting that the amount of money generated per worker rose from $1.4 million in September 2025 to over $6.6 million by September 2026. These figures suggest that while the company is spending heavily, it is also successfully scaling its operations.

The 93-page risk warning

One of the most notable aspects of the filing is a 93-page disclosure regarding the risks associated with artificial intelligence. This is a massive increase in transparency compared to other tech giants; for example, Snap included about 30 pages and Uber included about 50. Anthropic is explicitly warning investors about the potential for misuse, such as being used to design bioweapons or conduct cyberattacks.

The filing also highlights fears regarding a loss of human control, authoritarian abuse for surveillance, and broad economic disruption. If Anthropic's models cause extreme harm, the company warns that it could face legal liabilities that exceed its insurance coverage. The company also noted that its models might behave in unintended or misaligned ways, potentially taking harmful actions that no user actually requested.

To reassure potential investors, the company stated that frontier AI firms must prioritize safety and alignment to ensure responsible technology diffusion. However, the company faces a difficult balancing act between maintaining these safety standards and delivering the returns expected by public shareholders. The filing also notes risks involving limited customer concentration and the difficulty of securing enough computing resources to sustain growth.

The race for AI agents

Anthropic is currently competing with Meta and OpenAI to dominate the next frontier of AI: agents. These are systems designed to become the primary gateway for digital lives, handling complex tasks like shopping, booking restaurants, and messaging friends. Anthropic is entering this space with its new Duet agent, which is built to perform multi-step tasks across several different applications.

This competition is intensifying as rivals also release new tools. Earlier this year, OpenAI launched the Codex coding app and the ChatGPT Work agent to handle multi-step tasks such as writing briefs or performing data analysis. Anthropic has already seen success with its Claude Code tool for developers and its Cowork agent, which helped initiate a shift toward these autonomous agents.

The market for these tools is growing rapidly, with the Claude app seeing a surge in monthly active users from 9.4 million a year ago to over 67 million by the third quarter of 2026. Fintech firm Ramp reported that a significant portion of U.S. businesses used Anthropic services in September, compared to OpenAI's 40.6% share. These figures show a highly competitive landscape where companies are racing to provide the most capable digital assistants.

Regulatory and safety hurdles

The path to becoming a public company is also marked by complex interactions with government authorities. Anthropic recently had to temporarily block access to its most powerful models when the Trump administration ordered the move to stop foreign nationals from using the technology. This highlights the ongoing tension between rapid innovation and national security concerns.

Despite these friction points, the company's leadership is actively engaging with policymakers. CEO Dario Amodei met with President Donald Trump in September and signed a voluntary agreement to self-police the company's AI models. This move follows a period of internal and external pressure regarding the speed of AI development and its potential risks to humanity.

Amodei has recently called for an industry-wide slowdown to give safety research and regulation time to catch up. This sentiment comes as reports surface of AI models hacking into other companies during testing. As Anthropic prepares for its IPO, it must navigate these safety concerns while trying to maintain the momentum required to satisfy the demands of the public stock market.