3 Major Risks With Anthropic's IPO

Anthropic finally filed its long-anticipated prospectus last week, giving investors a deep look into the artificial intelligence (AI) company's finances, opportunities, and risks.

Management claims the company has a massive $40 trillion total addressable market, which includes the economic value of "knowledge work," the cumulative global impact of AI solutions and services, and other areas where AI is making an impact or could in the future.

The vast market opportunity and spectacular growth -- revenue increased 1,300% in the first half of the year over the first half of 2025 -- sound enticing from an investment perspective. However, there was a book-length section on the company's risks. While there are a lot of them, here are three big ones to note as you weigh up Anthropic's initial public offering.

Image source: Getty Images.

1. Anthropic has lots of competition

Anthropic is the creator of Claude, one of the most popular AI models. Management sees an eye-popping $40 trillion total addressable market (TAM), which means that it sees that number as the total possible amount of money it can make, and that's larger than the entire U.S. gross domestic product of $35.6 trillion today. Anthropic has $19.7 billion in trailing-12-month revenue today, and it is growing rapidly; second-quarter sales increased 1,368% year over year.

However, it's far from the only game in town. Other popular AI models, like Alphabet's Gemini and OpenAI's ChatGPT, are working hard and spending billions to stay competitive, and plenty of other companies are working on ways to make technology cheap or free.

When China's low-cost DeepSeek launched in 2023, it initially hit AI stocks hard. While they have mostly recovered, the threat of something like that happening again looms. Other companies, like Apple, are running on-device AI models that can process data locally. Since technology can evolve quickly, Anthropic may not always have the upper hand.

In a healthy market, competition works for the consumer and often brings prices down. Improved technology also brings prices down, and in order to stay competitive, Anthropic may have to do so as well at some point.

It may also not continue to grow as fast as it is right now for much longer, even at current pricing. Four-digit revenue increases are exciting, but they're not at all sustainable over the long term.

2. The price of much-needed compute power is likely to rise

Anthropic's rapid growth is driving profitability. It reported $974 million in operating income in the second quarter after losses, but that doesn't tell the whole story.

It has major contracts with many partners for years of compute power, which it needs for the training and inference that make its models work. "Our ability to develop, train, improve, and deploy frontier models depends on our access to enormous and increasing quantities of specialized compute," it says.

It's locked into long-term, multiyear agreements with Google, Amazon, Broadcom, Space Exploration Technologies, Microsoft, Advanced Micro Devices, and more. Under some current agreements, it has $265.6 billion in future commitments that range from one to 20 years. However, there are several more that don't specify the exact future commitment but are expected to be "significant." Depending on estimates, that could be more than $500 billion.

Management noted its efficiency, with revenue increasing 1,368% in the second quarter while compute and infrastructure expenses increased only 300%. I'd look out for those trends reversing, though.

3. Anthropic management retains voting rights

The company has an unusual capital structure that's not in shareholders' favor. The way it's presented is that the seven-member founding team, structured as a limited liability company (LLC), gets one share that carries 50.1% of the voting rights.

Management says that it believes that its "success depends in part on establishing a capital structure that enables our co-founders, management, and board of directors to focus on strategic initiatives that create lasting value for our stakeholders," instead of being forced to meet short-term goals. It also acknowledges that its unique structure "may effectively eliminate the ability of holders of our Class A common stock to influence important decisions, including a change in control."

These kinds of decisions becoming detrimental to investors may be unlikely, but investors should consider the implications.

Jennifer Saibil has positions in Apple. The Motley Fool has positions in and recommends Advanced Micro Devices, Alphabet, Amazon, Apple, Broadcom, and Microsoft. The Motley Fool has a disclosure policy.