Anthropic disclosed in its initial public offering prospectus that shifting government policies toward artificial intelligence could disrupt customer relationships and damage its reputation. The company, which is preparing for a potential IPO that could value it at about $2 trillion, warned that such actions could lead to lost revenue and operational interruptions.

Government encounters cited in prospectus

The filing detailed several encounters with the US government over the past year. In February, an order directed federal agencies to stop using Anthropic’s AI models. Subsequently, the Department of Defense designated the company a national security supply-chain risk.

In June, the Department of Commerce imposed worldwide export restrictions on two Anthropic models. Although Commerce later removed these restrictions and access was restored, Anthropic cautioned that comparable actions could recur.

Revenue exposure and risk disclosures

Anthropic stated that government contracts account for less than 1% of its annual revenue. Despite this low direct dependency, the prospectus acknowledged that advanced AI systems could pose "catastrophic or existential risks to humanity." The company noted that perceptions of these risks, influenced by government actions, could harm customers, partners, employees, and investors.

Trump administration stance shifts

The disclosure follows President Donald Trump’s softened stance toward Anthropic CEO Dario Amodei after an October 2026 dinner. Trump told Time that Amodei made a more favorable impression than before. The two met privately on Sunday, followed by a broader meeting with AI and tech leaders to discuss AI growth and societal impact.

Trump also suggested the government could potentially take an equity position in AI companies such as Anthropic and OpenAI, drawing comparisons with the government’s stake in Intel Corp (NASDAQ: INTC). However, he ruled out nationalizing top frontier AI labs.

What the numbers show

The prospectus reveals a divergence between Anthropic’s direct financial exposure to government contracts and its operational vulnerability to regulatory perception. While government contracts represent less than 1% of annual revenue, the company identifies government actions as a primary source of potential disruption to customer relationships and reputation. This suggests that regulatory risk is not merely a direct revenue issue but a broader market confidence challenge, particularly as the company seeks a valuation near $2 trillion.

Anthropic's initial public offering prospectus reveals that Broadcom Inc. (NASDAQ: AVGO) has agreed to provide the AI company with as much as $42 billion in lending tied to its long-term plan to secure massive amounts of computing capacity. This arrangement positions Broadcom not only as a chip supplier but also as a major source of funding for Anthropic's infrastructure buildout.

The disclosure, reported by Reuters, highlights a structural shift in Anthropic's financial dependencies. While previous filings highlighted sales concentration through Amazon and Google, this new data underscores Broadcom's dual role in hardware supply and financing. The prospectus notes that Broadcom's relationship with Anthropic stretches across hardware, leasing, and financing, distinguishing it from backers like Amazon that primarily contribute cloud reach and distribution for the Claude model.

Deal structure and financing terms

The IPO document outlined specific structural features of the lending arrangement. Broadcom retains the ability to name a financing partner, and the debt could potentially be converted into Anthropic equity. Anthropic stated it does not expect any notes to be sold before the IPO is completed. The convertible note could cover roughly one-third of Anthropic's committed five-year lease totaling $125.2 billion for tensor processing unit (TPU) computing capacity.

In April 2026, Anthropic placed cash into a restricted account for Broadcom's benefit and may have to add more funds under certain conditions. The company cautioned that some defaults could accelerate a large portion of lease payments while limiting access to the $42 billion facility for those obligations.

Strategic context and market reaction

Seaport Research analyst Jay Goldberg framed Broadcom's move as a competitive response to Nvidia's use of its balance sheet to support chip demand. "Nvidia is putting in place a massive amount of its balance sheet, and Broadcom is having to follow suit," Goldberg said. Rothschild & Co managing partner Robert Leitao questioned whether the revenue base can keep pace with the scale of financing, noting a concentrated bet on two companies generating enough revenue to support the financing.

Anthropic flagged potential conflicts of interest tied to relying on Broadcom for both hardware supply and financing. The company warned that Broadcom's pricing and hardware decisions could influence its ability to obtain enough compute. The filing indicated that Anthropic is expected to become Broadcom's biggest compute customer in 2027, with Broadcom projecting AI semiconductor revenue of about $115 billion in fiscal 2027 and $230 billion in fiscal 2028.

Existing financial landscape and risks

Prior disclosures in the same prospectus revealed that $2.16 billion of Anthropic's 2025 revenue came from sales through Amazon and Google, representing approximately 47% of total revenue. The company generated nearly $4.6 billion in total revenue in 2025 but recorded a net loss of $42 billion. As of December 31, the company held $20.28 billion in cash, cash equivalents, and short-term investments.

The technical issues arise as Anthropic prepares for an IPO targeting a valuation above $2 trillion. Robert Kindler, global chair of the M&A Group at Paul Weiss, does not expect the listing to move the broader stock market, noting that companies can now raise large amounts of private capital. Conversely, tech analyst Dan Ives has called the potential listing a watershed event for the technology sector.

What the numbers show

The combination of existing and new data reveals a complex dependency structure. Anthropic relies on Amazon and Google for distribution (47% of revenue) while simultaneously relying on Broadcom for capital expenditure financing ($42 billion) and hardware supply. With total future cloud and infrastructure spending planned at roughly $518 billion, the Broadcom facility covers a significant portion of near-term TPU leases. This creates a scenario where Anthropic's operational capacity and financial leverage are deeply intertwined with two key partners: one controlling market access and the other controlling both the supply chain and a substantial portion of the funding mechanism.

Global impact of recent disruptions

The service interruption affecting users internationally over the preceding 24 hours was resolved within approximately 40 minutes. User-reported outages on Downdetector peaked at 10,105 in the U.S. around 10:30 am ET. Anthropic's status page indicated elevated errors across claude.ai, Claude Code, Claude Cowork, and the Claude API starting at about 10:21 am ET. The company confirmed the issue was fully resolved by 10:59 am ET.