Summary of Key Points
Anthropic, an AI company, is expected to go public in October this year with a valuation of approximately $2 trillion. However, its repeated calls for AI regulation have raised questions: Is it truly aimed at protecting humanity, or is it more about creating barriers for itself (raising the entry barriers for the industry)? CEO Amodei responded publicly, emphasizing that reasonable regulation is necessary to curb the power of large companies and allow space for smaller competitors and open-source models. While the company has seen rapid growth in performance, it has not yet achieved net profit. Before going public, Anthropic faces challenges such as competition from Chinese open-source models and strained relations with governments; its high valuation must be supported by actual business performance.
Detailed Analysis
1. The Debate: Is the Call for Regulation About “Protecting Humanity” or “Building Barriers”?
The core issue is whether Anthropic, as a leading AI company, is using its emphasis on AI risks and calls for regulation as a means to create barriers to prevent smaller companies from entering the market and thus monopolize it.
The controversy was sparked by comments made in a podcast by investor Gavin Becker, who argued that Amodei’s constant talk about AI risks has contributed to increased opposition to AI in the United States (for example, through restrictions on data center construction). More importantly, there is a disagreement between the two regarding the approach to regulation: Becker advocates for “widespread distribution of AI capabilities” to spread the benefits and reduce risks, while Amodei believes that regulation is the key to balancing power concentration.
2. Amodei’s Response: Regulation as a “Fair Rule,” Not a “Barrier”
Amodei refuted the notion that regulation leads to power concentration, using a relatable example: A formal judicial system may be rigid, but it provides better protection for the vulnerable than private punishments. Similarly, regulation is not meant to favor large companies but to constrain their power. He cited two specific examples:
- The California SB53 law requires only large companies with annual revenues over $500 million to comply;
- Proposals for testing mechanisms require advanced models (owned by large companies) to undergo stricter security tests, while smaller companies’ models are exempted.
He also pointed out the “scaling law” of AI—larger models with more data have greater capabilities—and how open-source models can only partially mitigate this issue. Proper regulation can address three challenges: cyber/biological security risks, limiting the power of large companies, and providing opportunities for open-source models to develop. Additionally, he agreed with the regulatory direction set by the Trump administration, which requires testing of advanced models before deployment and also applies to open-source models if they are comparable in capability.
3. The “Basis” for the $2 Trillion Valuation and the “Concerns”
The Basis for the High Valuation:
- Anticipated annual revenue by the end of 2025 is around $9 billion, which soared to $47 billion in May 2026;
- Revenue in the second quarter of 2026 exceeded $11.5 billion, a 14-fold increase year-over-year;
- The company achieved its first operating profit (earnings after costs but before taxes, not net profit).
Concerns:
- The company has not yet reported net profit;
- Investors are concerned about three main issues: competition from Chinese open-source models (which are cost-effective and spread quickly), strained relations with the U.S. government (AI companies often face tensions over data security), and the potential for a similar stock price crash like that of SpaceX after its initial public offering.
4. The “Life or Death” Test of a High Valuation: More Than Just a Story
Amodei’s response focused mainly on regulation, which seems more like a defense of the company’s actions. However, a valuation of $2 trillion cannot be sustained solely by arguments about the need for AI regulation.
The AI industry is highly competitive, with model manufacturers updating their technologies almost weekly. A high valuation requires actual business achievements, such as converting technology into customer orders. Whether Anthropic truly deserves this valuation will depend on whether its models can remain leading and whether it can turn its technological advantages into stable net profit.
In summary, Anthropic’s high valuation is a double-edged sword: it reflects the market’s optimism about the future of AI but also comes with significant performance pressure. Behind the regulatory debate lies a power struggle over who will set the rules in the AI industry. In the end, only technology and business performance will determine who emerges as the winner.