Anthropic opens share sale to employees at $350 billion valuation

Three hundred fifty billion dollars. According to Bloomberg, that's what Anthropic is now worth. Just four years ago, it was a modest startup founded by a group of former OpenAI employees. Now it is offering current and former staff the chance to sell up to six billion dollars' worth of shares, at the same valuation set in its recent thirty-billion-dollar funding round. For dozens, and possibly hundreds, of engineers and researchers, that means one thing: they can turn lines in an option agreement into real money.

The mechanism Anthropic is using is familiar in Silicon Valley. A secondary share sale, known as a tender offer, lets employees of private companies sell their stakes without waiting for an IPO. Buyers are usually institutional investors, hedge funds, and dedicated secondary-market funds willing to pay a premium for access to the hottest private assets. At $350 billion, Anthropic's valuation puts it in the same league as giants like Salesforce or Netflix by market capitalization, even though it has never gone public.

You can't understand this deal without looking at the AI industry's arms race. Anthropic recently closed a massive $30 billion round, one of the largest venture investments in the history of the tech sector. The company behind the Claude model family is aggressively building out compute capacity, expanding its research team, and fighting for enterprise customers against OpenAI, Google DeepMind, and fast-rising Chinese competitors. The $350 billion valuation reflects current revenue less than it reflects investors' bet that Anthropic will end up among the two or three companies that shape the industry's future, though by all indications that revenue is growing fast.

For employees, the secondary sale does more than offer a payday. It's a retention tool. The AI industry is in an unprecedented war for talent: top researchers and engineers field offers worth tens of millions of dollars, and moves between rival labs have become routine. When employees can cash out part of their options now, instead of waiting for a hypothetical IPO years down the line, they have one less reason to leave for a competitor offering cash. OpenAI has used the same strategy repeatedly, running tender offers at rising valuations. Anthropic is essentially following that playbook, but the scale stands out: six billion dollars on the secondary market for a company not yet five years old.

The broader picture matters too. Valuations of leading AI startups have reached levels that seemed fantastical just two years ago. OpenAI is reportedly valued in the $300–350 billion range, and Elon Musk's xAI has raised money at a valuation above $50 billion. Together, these companies form a new class of "super-unicorns" with a combined valuation of more than a trillion dollars, and none of them has gone public yet. Critics point to signs of a bubble: AI companies' revenue is growing fast but still doesn't come close to justifying their valuations. Supporters counter that this is a technology that will reshape the global economy, and that current multiples reflect potential rather than the present state of the business.

The share sale also sends a reputational signal for Anthropic. By offering employees liquidity at its last-round valuation, the company shows confidence that the valuation is fair and durable. If leadership expected a correction, it would make more sense to delay the tender offer or run it at a more conservative price. The fact that outside investors are willing to buy employee shares on the same terms as the main round points to strong demand for a stake in the company.

Ultimately, Anthropic's story shows how quickly AI is rewriting the rules of venture capital. Founded by Dario and Daniela Amodei in 2021 with a mission to build safe AI, the company has become one of the most valuable private tech companies in the world in just a few years. The six billion dollars its employees can now take home is a generous bonus, but it also marks an era in which building language models generates wealth at a speed unseen even by Silicon Valley standards.