VCs Break Taboo by Backing Both Anthropic, OpenAI in AI Battle

(Bloomberg) — OpenAI and Anthropic are working to close two of the largest funding rounds in history and increasingly leaning on an overlapping pool of investors to meet their immense capital needs — a practice that was once taboo in startup land.

Most Read from Bloomberg

Sequoia Capital and Altimeter Capital, for example, are both expected to invest in Anthropic’s latest financing, which is set to bring in more than $20 billion. Sequoia first backed OpenAI in 2021, and has invested in several of the company’s rounds since. Altimeter, which has said that OpenAI was its biggest bet ever, is set to invest more than $200 million in Anthropic’s latest deal, according to a person familiar with the matter who asked not to be identified discussing private information.

The growing trend highlights how much the top few AI developers are scrambling the venture funding landscape. Silicon Valley investors have historically gone to great lengths not to bet on rival startups — choosing instead to pick a winner and back them to the end, avoiding perceptions of a conflict of interest. But with artificial intelligence breakthroughs creating a feeding frenzy for tech investors, those old rules are increasingly defunct.

“When you talk about OpenAI and Anthropic, these are generational companies, the likes of which we may not see again in our lifetimes,” said Ethan Choi, a partner at OpenAI-backer Khosla Ventures. “In the face of that, some VC firms have lowered the bar for conflicts.”

Some of the AI industry’s biggest backers are betting on both OpenAI and Anthropic. Tech giants Amazon.com Inc., Nvidia Corp. and Microsoft Corp. have all discussed plans to invest in both companies, Bloomberg has reported. Blackstone Inc., the world’s largest alternative asset manager, is in talks to increase its stake in Anthropic to $1 billion, after investing in OpenAI. And Abu Dhabi’s MGX, an OpenAI investor and partner on the startup’s Stargate data center project, is in talks to invest hundreds of millions in Anthropic’s latest funding round, pulling off the hat trick of investing in three of the largest model makers, including Elon Musk’s xAI.

Those investors join JPMorgan Chase & Co.’s growth equity investing arm, which holds stakes in OpenAI and Anthropic, as well as Iconiq, an Anthropic investor that gained a stake in OpenAI after the AI giant acquired one of its portfolio companies, Statsig.

The bets are the latest example of changing norms around investing in rival companies in recent years. As AI catapults startup valuations to new heights, mega-firms can’t seem to stay away from model makers. Famed venture firm Andreessen Horowitz has invested in both xAI and OpenAI, while early-stage-focused shop SV Angel, founded by Ron Conway, has made smaller investments in OpenAI and Anthropic.

Sequoia, JPMorgan, Andreessen, Iconiq and SV Angel declined to comment on their AI portfolios. Altimeter did not respond to requests for comment. OpenAI and Anthropic declined to comment.

Overlapping VCs can cause problems for AI startups, who fear that backers of competing companies will share proprietary information, or may otherwise compromise one of their investments in favor of another.

“The real risk is information leakage,” said Michael Torosian, a partner at the law firm Baker Botts LLP. “Founders just have to know that there is a true structural risk that they have to consider. And they have to think on the front-end about the diligence they have to do on these investors.”

For some VCs, like angel investors, their stakes are usually too small to get information rights at big companies. OpenAI, for example, shares certain financial data only with backers who invest above a certain amount in the company, according to a person familiar with the matter. It’s usually when firms have larger stakes in competing companies that conflicts can arise.

Investors will sometimes try to solve for this problem by asking separate partners to be involved in different investments — and putting up an “invisible wall” between them, said Deedy Das, a partner at Anthropic investor Menlo Ventures. Still, knowing that a firm is backing multiple players can make founders nervous. Some startups have even discussed enforcing non-disclosure agreements or alternative legal safeguards, Torosian said.

Other firms, such as Khosla Ventures and Thrive Capital, have refrained from backing Anthropic. Vinod Khosla, founder of Khosla Ventures, was one of OpenAI’s earliest backers. At Khosla, growth investor Choi told Bloomberg he’s a “one-woman and one-LLM man, as is Vinod.”

Josh Kushner, founder of OpenAI investor Thrive Capital, echoed that sentiment in a post on X last year, responding to a chart pointing out overlapping AI investments in Silicon Valley. “Call us old school,” he wrote. “But we are serial monogamists.”

OpenAI and Anthropic have already tapped many of the leading venture funds and Big Tech firms to raise the money needed to fund chips, data centers and talent — and their capital needs only appear to be growing. That could make it harder to turn down checks from overlapping backers. OpenAI, for example, is looking to raise a record $100 billion.

At the same time, as VC firms get larger, and the appetite for AI stakes grows, it’s more difficult for investors to turn down competing investments. If “they’re such a big fund, and they’re so interested in the space, it feels inevitable,” Torosian said.

In a post on X in 2024, Altimeter founder Brad Gerstner dismissed fears around investor double dipping. Great founders and startups “win or lose on the field,” he said, adding that Altimeter is a “happy investor” in OpenAI, Perplexity, Meta and Google, among other competing companies.

Many tech investors believe that betting on both OpenAI and Anthropic does not pose a conflict of interest because OpenAI’s ChatGPT is in large part a consumer-facing product, while Anthropic has targeted developers. However, OpenAI has recently been promoting a tool for developers, even as Anthropic’s Claude has in some ways become a breakout consumer hit.

As AI moves into more traditional business areas, some investors are backing multiple smaller startups tackling the same industry. For example, Khosla, the major OpenAI backer, has invested in multiple coding-related AI startups including Cognition, Replit, Lovable and Emergent. In an emailed statement, Vinod Khosla said, “In all cases we view them as going after very different markets.”

In legal tech, Sequoia has backed both Harvey, last valued at $8 billion, as well as other legal AI companies Crosby and Sandstone. Crosby co-founder Ryan Daniels said he doesn’t consider his startup a direct competitor to Harvey, exactly. Daniels’ gauge for “competitor” boils down to: If you buy services from one company, do you not need the other one? “That definition didn’t apply to us and Harvey.”

In the health care sector, Andreessen Horowitz invested in AI medical scribe startup Abridge in June at a $5.3 billion valuation, several years after backing its competitor, Ambience Healthcare. The bet caused a stir within the firm, since Andreessen Horowitz’s growth fund made the Abridge bet without the health care fund’s prior approval, three people with knowledge of the situation said.

The firm’s co-founder Ben Horowitz later met with Ambience to reassure the company’s leadership about its commitment, two people said — and Andreessen Horowitz went on to co-lead Ambience’s next fundraise.

For many investors, backing multiple AI players is worth the risk that company founders will object. “If you have a winner in your portfolio and you try to invest in a competitor, you risk ruining your relationship with that winner,” said Menlo Ventures’ Das. “The tradeoff is, of course, you might get great returns.”

–With assistance from Dina Bass and Shirin Ghaffary.

Most Read from Bloomberg Businessweek

©2026 Bloomberg L.P.