OpenAI’s skyrocketing valuation naturally invites doubt. The company’s valuation surged from roughly $30B in 2022 to about $500B this fall and is expected to reach $830B by the end of March (WSJ), representing roughly a 27x increase over four years. This jump has coincided with feverish excitement around AI, along with rising investor concerns that expectations (and valuations) may be running well ahead of reality.

Another red flag is the level of hype surrounding AI. We’ve seen venture investors, the media, and corporate messaging all proclaim AI as the next revolution. We see it in our venture practice, with early-stage valuations on average 2.5x higher today than three years ago. With all of the mark ups, its making it harder to separate long term winners from fad chasers.

OpenAI’s financial outlook also gives pause. Based on rumored reports of internal forecasts, OpenAI will likely lose between $100-$150B from 2025-2029. That said, in the coming years the company should pull in enough cash from investors like Nvidia and others to cover the losses. These steep losses reflect huge spending on computing infrastructure and talent. For an investor, funding a business that might not break even until the end of the decade, if all goes well, is a daunting prospect.

History provides further reason for caution. The late ’90s tech boom was filled with apparent winners that ultimately disappeared. Netscape Navigator once dominated web browsing; Lycos was a leading search portal; Pets.com symbolized the online buying frenzy. Yet two decades later, none of these darlings survived. The lesson is that being early in a transformative tech wave is no guarantee of long term victory, and the hype can propel early “winners” that later fade away.

Finally, the war for talent in AI is intensifying and could undermine OpenAI’s edge. The company’s leadership in generative AI makes its top researchers and engineers prime targets for deep-pocketed competitors and the lure of jumping ship to start the next big thing. We’re all aware of how much Meta is spending on its superintelligence lab to bring in AI experts. As more companies (from startups to giants like Google and new AI-first companies like xAI) seek talent, salaries and stock grants will continue to skyrocket. OpenAI may struggle to retain its best people, which feeds back into the high-cost, high-burn concern. The bottom line is the bear case has merit: a big valuation, AI hype, big losses, echoes of a tech bubble, and fierce competition for the people who drive innovation.

It’s important to consider that many of the world’s most capable tech leaders are all in on AI’s potential. Huang, Nadella, Pichai, Zuckerberg, and Musk are effectively betting their companies on AI. They are reallocating tens of billions of dollars of resources, shifting product strategies, and making bold statements about AI’s potential. This collective conviction from industry veterans lends credibility to the idea that AI’s substance will eventually catch up to, or exceed, the current hype.

These CEOs have been unabashedly optimistic in public forums. Microsoft’s Nadella has said that AI will help solve some of the world’s most complex problems. Zuckerberg recently urged people to “take superintelligence seriously,” calling AI “the most important technology of our lifetimes.” He emphasized that AI should serve humanity rather than just sit in a data center automating tasks across society. When leaders of this caliber are essentially staking their companies’ futures on AI, it suggests they see real transformative power beyond the buzz. They’ve witnessed multiple tech cycles (mobile, cloud, internet) and succeeded by embracing these shifts.

One encouraging factor is that today’s AI reality is far more modest than those grand visions, underscoring how early we still are. So far, AI is great at narrow tasks: engaging in chat conversations, generating images from prompts, writing code snippets, and handling basic customer service queries. These are impressive achievements, but relatively straightforward problems compared to reimagining healthcare. If the tech follows a trajectory similar to past breakthroughs, AI systems will progressively tackle more complex, real-world challenges. The leap from chatbots to curing diseases is enormous, but sustained investment and rapid progress should close that gap over time.

We believe the bullish outlook deserves serious consideration. In prior eras, when multiple competitors all committed to a technology (for example, mobile computing or cloud infrastructure), that was a strong signal the trend was more than hype. It became a self-fulfilling prophecy as the collective effort accelerated the shift. Similarly, if leaders like Nadella, Zuckerberg, Musk, and Huang are correct about AI, then its real world impact will only grow.

Setting aside emotion, how does OpenAI’s $830B valuation stack up on based on the financial outlook? The answer is it’s more expensive, but factoring in its growth 100% compared to other tech growing around 15%, the valuation is essentially inline with the public peers.

1. OpenAI’s implied 2026 revenue multiple vs. other high-growth tech companies. If OpenAI reaches about $35B in revenue in 2026 (midpoint of company forecast), a $830B valuation today equates to roughly 24× next year’s sales. That figure initially sounds high, but it’s in the ballpark of other fast growing software and AI peers.

For context, CrowdStrike, currently trades near 21× CY26 revenue and is growing around 22%. Other cloud software names like Shopify and ServiceNow have revenue multiples in the mid teens (about 12–16× sales) while growing roughly 14-47% annually. Even chipmakers tied to the AI boom can command elevated multiples: Broadcom has seen multiples in the high single-digits while delivering ~20% growth. By these standards, OpenAI at 24× forward sales is only slightly above the peer group median.

2. OpenAI’s implied 2027 revenue multiple vs. other high-growth tech companies: Where OpenAI stands out is its growth rate. We estimate OpenAI’s revenue is on track to increase from about $4B in 2024 to about $15B in 2025, and then to $35B in 2026, and $70B in 2027. No public tech company at $35B scale is growing anywhere near that fast. The group of high-growth companies mentioned above (CrowdStrike, Shopify, ServiceNow) are all expected to grow next year on the order of 15% to 30%, which is impressive for their size but pales in comparison to OpenAI’s 100% growth. From that lens, OpenAI’s premium valuation seems rational, given it’s growing far faster than almost any comparable company.

3. OpenAI compares to the Mag 7 multiples for 2026 & 2027: We must also acknowledge that compared to the tech mega caps, which OpenAI will enter the public markets as a mega cap, its valuation is less attractive. The Mag 7 are currently trading at 9x CY26 revenue, compared to OpenAI at 24x. OpenAI is growing at 100% compared to the Mag 7 at 16%. If we fast forward to 2027 multiples, Open AI is trading at 12x compared to the Mag 7 at 8x.

The bottom line: We believe OpenAI can grow around 100% annually over the next several years. If that proves true, a $830B valuation looks objectively cheap given its growth rate. If it doesn’t, the valuation is going lower.

One argument in favor of OpenAI is the company’s user traction and pace of innovation. In October, Altman announced that ChatGPT now exceeds 800m weekly active users. We estimate that today it’s closer to 900m, which means it’s now used by about 17% of the world’s adult population (18-80 years old) every week. This kind of network scale provides a distribution advantage. OpenAI today can roll out new features and improvements at a reach that only Google and Meta can match. The key question is, can OpenAI hold onto this lead?

The encouraging sign is that OpenAI is not standing still. Over the past three months, they have rolled out the following updates:

• OpenAI is showing an ability to continue to frequently update its models. After the release of GPT-4 in early 2023, they came out with GPT-4.5 in 2024, and by August released GPT-5, GPT-5 Pro for developers and GPT-5.2 this month. They’re also releasing smaller specialized models like the GPT-Image-1-Mini model which tokens cost about 70% less than the premium models. In effect, they are running two races at once: one to grow adoption (user base and integrations) and another to advance the technology itself.
• Disney content agreement for Sora.
• Pulse, a new feature for Pro users that proactively delivers personalized daily AI curated updates. Instead of waiting for users to prompt it, ChatGPT can now research and push information to users each morning, increasing engagement. This plays into a long-held Deepwater thesis we call the “anticipation economy”. Then at DevDay in October, OpenAI Apps in ChatGPT, which means ChatGPT is becoming a platform or an app ecosystem of its own, like an AI app store, where users can seamlessly access outside services (from education to shopping) without leaving the chat. This hooks users into using ChatGPT for daily tasks.
• AgentKit, a toolkit for developers to build agents. These agents can take on multi-step tasks and operate across other software and services. This goes well beyond ChatGPT that simply answers questions.
• An upgraded Codex (its AI coding assistant). Codex can integrate directly into tools like GitHub, allowing it to write, test, and even self-correct code with minimal human.
• Atlas browser for MacOS.

Disclaimer