I remember the first time I saw a revenue projection that made my data scientist instincts scream. It was 2016, and a startup claimed it would capture 30% of the global remittance market within three years. I ran the numbers—back-of-the-envelope, based on total addressable market and realistic adoption curves—and the math simply didn't work. The company folded within eighteen months, but not before it had raised millions from investors who mistook a good story for a good model.
Today, I see the same pattern playing out in the AI industry, and it carries profound implications for the blockchain world. A recent analysis of Anthropic's IPO valuation suggests the company could hit $190-200 billion in revenue by 2028. That's not a typo—or maybe it is. Either way, as a decentralized protocol PM who has spent years building trust through transparency, I can't ignore the structural flaws in this narrative. And for those of us building on-chain AI, this is a cautionary tale about the dangers of centralized valuation storytelling.
Context: The Parable of the Overconfident Prophet
Anthropic, the AI safety company behind the Claude model family, has become a darling of Silicon Valley. Its valuation has soared past $60 billion, and whispers of an IPO in 2026 or 2027 have analysts scrambling to justify the numbers. The revenue projection in question—$190-200 billion by 2028—appears in a document that circulated among investors. It claims that Anthropic will capture a massive share of the enterprise AI market, driven by the adoption of its Constitutional AI approach and partnerships with Amazon and Google.
But here's the thing: I've spent the last decade analyzing protocol economics and revenue models, first as a data scientist in Buenos Aires and later as a PM for decentralized lending protocols. I've seen what happens when projections ignore the fundamental constraints of market size, competition, and adoption velocity. The $190-200 billion figure doesn't just stretch credibility—it breaks it.
To understand why, we need to look at the numbers through the lens of blockchain thinking: decentralization of trust, community-driven verification, and the unyielding logic of supply and demand. If Anthropic were a protocol, I'd be running a governance proposal to audit its revenue assumptions before allocating any capital.
Core: The Numbers Don't Lie—But People Do
Let's start with the arithmetic. Anthropic's current annualized revenue is estimated at around $1 billion (some sources say $1.5 billion, but let's be conservative). To reach $200 billion in four years, the company would need a compound annual growth rate (CAGR) of approximately 280%. That means every year, Anthropic must nearly quadruple its revenue.
I've modeled this scenario for a dozen DeFi protocols—Aave, Compound, Uniswap—when they presented their revenue forecasts. The mathematical reality is that such growth rates are only sustainable for one or two years at most, typically from a very small base. For a company already at $1 billion, the base is too large. The raw numbers:
- Year 0 (2024): $1B
- Year 1 (2025): $4B (4x growth)
- Year 2 (2026): $16B (4x)
- Year 3 (2027): $64B (4x)
- Year 4 (2028): $256B (4x)
That's $256 billion, not $200 billion. But even $200 billion requires a 200x growth over four years. To put that in perspective, the entire global cloud infrastructure market was about $350 billion in 2024. Amazon Web Services, the dominant player, generated $105 billion in revenue after 12 years of aggressive expansion. Anthropic is supposed to surpass AWS in four years? Not from a single product line selling API access and enterprise subscriptions.
I've built a sensitivity table to show the range of possible outcomes based on realistic growth rates:
| Growth Scenario | 2025 Revenue | 2026 Revenue | 2027 Revenue | 2028 Revenue | Achieves $200B? | |----------------|--------------|--------------|--------------|--------------|-----------------| | Conservative (100% CAGR) | $2B | $4B | $8B | $16B | No | | Moderate (150% CAGR) | $2.5B | $6.25B | $15.6B | $39B | No | | Aggressive (200% CAGR) | $3B | $9B | $27B | $81B | No | | Extreme (300% CAGR) | $4B | $16B | $64B | $256B | Yes (barely) |
Only the extreme scenario—which requires four consecutive years of tripling revenue—gets anywhere close. And that scenario assumes no market saturation, no competition, no regulatory hurdles, no macroeconomic downturn. It assumes Anthropic captures the entire growth of the AI market while everyone else stagnates.
But here's the blockchain-inspired insight: the market is not a deterministic machine. It's a network of interdependent agents. In a decentralized system, no single entity can capture all value because value flows to the most efficient node. Anthropic is a centralized node. Its revenue projections are essentially a claim that it will be the most efficient AI provider by a factor of ten over its nearest competitor, OpenAI.
Let's compare with OpenAI's own projections. Wall Street estimates OpenAI's 2028 revenue at around $100 billion, half of Anthropic's claimed figure. OpenAI has a first-mover advantage, a massive user base (ChatGPT alone has 200 million weekly active users), and a partnership with Microsoft that gives it preferential access to Azure infrastructure. If Anthropic were to surpass OpenAI by a factor of two, it would need to capture market share from the dominant player while also expanding the total addressable market. That's possible in theory, but the numbers don't support it.
I've seen this dynamic before in DeFi. During the 2021 bull run, every lending protocol projected exponential growth. Compound's governance token was trading at $800, implying a market cap of $8 billion, while its total value locked was only $10 billion. The revenue projections assumed that DeFi would replace traditional finance within three years. But the reality was slower adoption, competition from Aave, and regulatory uncertainty. Compound's token is now around $50. The lesson: projections that ignore competitive dynamics are not predictions—they are wishful thinking.
Contrarian: What If the Numbers Are Actually Right?
Let me play devil's advocate for a moment. What if Anthropic's $200 billion revenue forecast is not a mistake, but a deliberate signal? What if it's a narrative designed to justify a pre-IPO valuation of $1 trillion or more? In the world of crypto, we call this a "valuation game." Projects often present aggressive revenue projections to attract early investors, then adjust them downward after the funding round closes. The same mechanism applies in traditional venture capital, but with higher stakes.
There's another possibility: the revenue figure might include "pass-through" revenue from AI computing resale. If Anthropic sells API access that includes the cost of cloud compute from Amazon or Google, and then marks it up, the reported revenue could be inflated. This is similar to how some blockchains report "total value settled" instead of "fee revenue." In Ethereum, the actual revenue (transaction fees) is a fraction of the total value transferred. If Anthropic reports gross revenue including compute costs, then a $200 billion top line could correspond to $40-50 billion in net revenue—still high, but more plausible.
But even that requires a heroic assumption: that Anthropic can maintain its competitive edge. The AI landscape is shifting rapidly. The barrier to entry for new models is lowering, thanks to open-source alternatives like Llama and Mistral. In the blockchain world, we've seen how open-source protocols can erode the moat of proprietary platforms. Uniswap's open-source model allowed clones to proliferate, but the network effect of liquidity kept Uniswap dominant. For AI, the network effect is data and compute, not liquidity. Anthropic has a head start, but it's not insurmountable.
I've been in meetings where teams presented "Total Addressable Market" analyses that conveniently ignored the existence of competitors. In 2022, a DeFi project claimed it would capture 10% of the global stablecoin market within two years, ignoring the dominance of USDT and USDC. The project failed to reach even 0.1%. The same pattern appears here: the $200 billion figure assumes that Anthropic can capture 40% of the global AI software market, which is projected to be around $500 billion in 2028. That's a concentration that centralizes power in a way that contradicts the ethos of decentralization.
Takeaway: The Lesson for Crypto AI
As someone who has spent years advocating for transparent, community-driven valuation, I see Anthropic's projection as a warning. The blockchain industry is increasingly converging with AI, with projects like Bittensor, Render Network, and Akash Network offering decentralized alternatives to centralized AI infrastructure. These projects often face the same temptation: overpromise to attract capital.
But the crypto community has a built-in immune system: on-chain data, open governance, and the ability to verify claims independently. When a DeFi protocol claims a certain TVL, we can check it on Etherscan. When a Layer2 project claims a certain throughput, we can run our own tests. The same should apply to revenue projections. If Anthropic were a protocol, I would demand a revenue breakdown, an audit of its API usage, and a transparent model of its growth assumptions.
Until then, treat the $200 billion figure as what it likely is: a narrative tool, not a forecast. The real question is whether investors will demand evidence before buying into the story. In blockchain, we've learned that trust is earned through transparency, not built on projections. Connect first, transact second. Always.
Based on the data, I'd put the realistic range for Anthropic's 2028 revenue at $20-50 billion, not $200 billion. That's still a remarkable achievement—it would make Anthropic one of the fastest-growing software companies in history. But it's a far cry from the techno-utopian fantasy of dominating the world's AI infrastructure. For those of us building in the blockchain space, the lesson is clear: don't let the hype of centralized AI distract you from the hard work of building decentralized alternatives. The market will reward the projects that deliver real value, not the ones that tell the best story.