On December 12, Anthropic announced the completion of a $30 billion Series G funding round, bringing its valuation to $380 billion. The round was co-led by Singapore's sovereign wealth fund GIC and Coatue Management, with other investors including DE Shaw, Dragoneer, Peter Thiel's Founders Fund, and Abu Dhabi's MGX Fund. Microsoft and Nvidia also participated, contributing a portion of their previously committed $15 billion.
This is the largest funding deal so far in 2026 and the second largest venture capital funding round in history, second only to competitor OpenAI's $40 billion in 2025.
The list of investors reveals a reality: more than 30 institutions, from Sequoia Capital to Lightspeed Venture Partners, from Goldman Sachs to Morgan Stanley, from Blackstone to BlackRock, participated in this round. Sequoia Capital, in particular, holds shares in three companies: OpenAI, xAI, and Anthropic.
In Silicon Valley, investing in direct competitors in the same industry used to be an inviolable red line, but that line has been crushed in the AI era.
An unwritten rule
In Silicon Valley's venture capital world, there's an unwritten rule that's been around for forty years: don't invest in competitors.
The logic is simple. When you invest in a company, you're committing not just capital, but also trust. You'll sit on the board, see trade secrets, product roadmaps, customer data, and financial figures. If you've also invested in its direct competitor, how can you prove you haven't passed on A's intelligence to B?
This is not just a moral issue, but a matter of business reputation. In an industry that operates on word of mouth, the label of "betraying the founder's trust" is more fatal than a failed investment.
This is why Vinod Khosla, founder of Khosla Ventures, publicly stated in 2025 that he "would not invest in directly competing AI companies at the same time." Thrive Capital also chose loyalty: all-in on OpenAI, rejecting the temptation of other large AI models.
But Sequoia doesn't think so.
In late 2024, Sequoia Capital underwent a generational shift. Roelof Botha, who had long led the firm, stepped down as Global Managing Partner, and Pat Grady and Alfred Lin took over. The new leadership team made a radical decision: to simultaneously bet on three leading AI companies. Sequoia held an early stake in OpenAI, later invested in Musk's xAI, and now appears on Anthropic's investor list.
It's not just Sequoia. Altimeter Capital has invested over $200 million in Anthropic and also holds shares in OpenAI. Blackstone has invested approximately $1 billion in Anthropic. Abu Dhabi-based MGX Fund has invested in both OpenAI and Anthropic.
The smartest money in Silicon Valley is buying up every single horse on the track at the same time.
The speed at which money is burned
Why are investors willing to break taboos? Because AI is an arms race that no one can afford to lose. And the first rule of an arms race is: you can't stop.
Anthropic's annualized revenue (ARR) has climbed to $14 billion, maintaining a more than tenfold annual growth rate for three consecutive years. The number of enterprise customers with annual payments exceeding $1 million has increased from 12 to over 500 in two years, and the company projects that its ARR will surpass $30 billion by the end of 2026.
The most notable growth engine is Claude Code: a program assistant that allows writing and debugging code with minimal human intervention. This product has generated $2.5 billion in annualized revenue, doubling since the beginning of this year, with enterprise clients contributing more than half. Currently, 4% of public code commits on GitHub are completed using Claude Code.
Anthropic CFO Krishna Rao said in an official statement:
Whether it's entrepreneurs, startups, or the world's largest corporations, the message from our clients is the same: Claude is increasingly becoming a more critical part of how businesses operate. This funding round reflects the tremendous demand we've seen from our clients, and we will use this investment to continue building the enterprise-grade products and models they trust.
Recently, Anthropic's technology has also been shaking up the financial markets. Earlier this month, the company quietly released a tool that automates certain legal tasks, triggering a chain reaction of declines in legal services stocks. It then launched a new AI model optimized for enterprise tasks, including financial research, causing financial services company stocks to fall.
But income is only half the story; the other half is expenses.
In 2025, Anthropic spent $2.66 billion on computing resources on AWS alone. Adding in researcher salaries, data procurement, and GPU cluster construction, Anthropic's total annual expenses far exceeded its revenue. The company expects to reach break-even no earlier than 2028.
In layman's terms, this is a company with annualized revenue of $14 billion, but it's still burning through cash. It needs to continuously raise funds, not because it's unsuccessful, but because the cost of success is outpacing revenue growth.
This is the harsh reality of large-scale AI model commercialization. Your revenue can grow like a rocket, but your computing costs will grow even faster ; the training cost of each generation of cutting-edge models is 3 to 5 times that of the previous generation.
Anthropic has announced a $50 billion investment in data centers across the United States, with facilities in Texas and New York expected to open this year. The company also plans to use Google's multi-billion dollar specialized AI chips. However, these investments pale in comparison to OpenAI's commitment to investing over $1.4 trillion in AI infrastructure over the next few years, while also seeking to raise up to $100 billion in new funding.
This explains the necessity of the $30 billion in funding. Anthropic isn't raising "growth money"; it's buying its survival.
The logic of fear of missing out
So, let's get back to the core question: Why are investors willing to bet on multiple AI companies at the same time, even at the cost of breaking a 40-year-old industry taboo?
The answer lies in a deeper fear.
By 2025, global investment in AI is projected to exceed $150 billion. However, this funding is highly concentrated, flowing to fewer than five companies: OpenAI, Anthropic, xAI, Google DeepMind, and MetaAI. The entry ticket to this race is so expensive that only sovereign wealth funds and top-tier venture capital firms can afford it.
In this environment, the cost of missing out on the winner is far greater than the loss of betting on the wrong horse.
Imagine you're Sequoia Capital. If you only invested in OpenAI, and Anthropic ultimately wins, you not only lose the returns from Anthropic, but you'll also be remembered as "the fund that missed out on the biggest winner of the AI era." In the venture capital industry, reputation is more valuable than a single return. A fund that missed out on Google, and a fund that invested in both Google and Yahoo—the latter will be remembered much longer.
Therefore, hedging is not a strategy, but insurance.
But here's the paradox. When all the smart money is hedging the same bet, what they're actually doing isn't diversifying risk, but turning the entire AI industry into a giant pool of funds. Regardless of which company ultimately wins, capital can ensure it's on the winner's side.
Those who cannot participate in this hedging game—small VCs, individual investors, and ordinary employees—are excluded. They can only choose one side and then wait.
The company of the departed
To understand Anthropic today, let's go back to a resignation in December 2020.
Dario Amodei was formerly the Vice President of Research at OpenAI. During his four years at OpenAI, he led the development of GPT-2 and GPT-3, two models that changed the trajectory of the entire AI industry. When he joined, OpenAI was still a non-profit research lab. When he left, it had become a commercial company 49% owned by Microsoft.
In late 2020, Dalio and his sister Daniela Amodei jointly submitted their resignations. According to multiple sources familiar with the matter, the core of the disagreement was a conflict between security and commercialization. Dalio believed that as model capabilities rapidly improved, OpenAI's investment and decision-making power in security research were being gradually diluted. Microsoft's multi-billion dollar investment accelerated this trend.
In other words, when your biggest investor says "get the product out quickly," the voice of security researchers is relegated to the background.
In January 2021, Dalio founded Anthropic with seven core researchers from OpenAI. Their mission was clear: to build a "responsible AI company" that strikes a balance between commercial success and AI safety. The company's name comes from the Greek word "anthropos," meaning "human": a choice imbued with a certain idealism.
Looking back five years later, the scale of this economic exodus is astonishing. In May 2021, it raised $124 million in Series A funding. In 2023, Google led the investment, bringing the valuation to $4.1 billion. In 2024, Amazon increased its investment, pushing the valuation above $18 billion. In March 2025, it reached $61.5 billion. In September of the same year, it reached $183 billion.
Then comes the $380 billion in February 2026. Having just raised $13 billion a few months prior, this latest round nearly doubled its valuation. Anthropic also announced that it will allow employees to sell their shares at the valuation of this funding round.
In five years, Anthropic transformed from a security research lab into one of the world's most valuable AI companies and the fourth most valuable privately held company globally, raising nearly $64 billion in funding. The seven people Dalio took with him now support a company with over 1,500 employees.
After the taboo disappeared
Ironically, what underpins everything behind Anthropic isn't a security narrative, but rather the logic of an arms race. Investors are betting on Anthropic mostly not because it's safer, but because they can't afford the consequences of not being present.
In the age of AI, loyalty is a luxury. A 40-year-old unwritten rule in Silicon Valley—not investing in competitors—exists because most markets have enough time for winners and losers to naturally differentiate themselves. You can spend five or ten years observing a sector before deciding where to place your bet.
But AI is different. The window of opportunity for this competition is too short, the stakes too high, and the number of participants too small. Under these conditions, hedging is not betrayal, but rationality. And when everyone chooses rationality, the taboo is no longer taboo. It is simply a line that everyone tacitly crosses.
Because in Silicon Valley, the real taboo is never investing in competitors. It's missing out on the next era.
