Anthropic's $65 Billion Run Rate: A Victory for Centralized AI, Not for Decentralization

We didn't see this coming. Anthropic, the AI company that built its brand on ethical alignment and safety, is now racing past OpenAI with a $65 billion annualized revenue run rate as of July. That's $25 billion ahead of its rival, according to people familiar with the figures. The company plans to go public as soon as this fall, with investors expecting a valuation north of $2 trillion. But as someone who has spent years auditing token distributions and advocating for transparent governance, I can't help but ask: are we celebrating the wrong kind of growth? Let's start with the numbers. Anthropic's annualized run rate expanded 622% since late 2025, from roughly $9 billion at year-end to $47 billion in May, then to $65 billion by July. The May-to-July jump alone added $18 billion, a 38% gain in two months. Preliminary second-quarter revenue topped $11.5 billion, up from $787 million a year earlier. Adjusted operating income turned positive. On the surface, this is a stunning financial trajectory. But run rate is a dangerous metric—it annualizes a single month's performance. If Anthropic's July revenue was $5.4 billion, the run rate assumes that pace holds for 12 months. That's a heroic assumption for a company that relies on enterprise contracts, API usage, and subscription renewals that can fluctuate wildly. We didn't learn this lesson from the 2017 ICO boom? I led a volunteer audit team back then for a prominent Ethereum-based utility token. The project's whitepaper promised decentralization, but the token distribution favored insiders. I published a critique that forced the team to revise its allocation. The lesson was clear: transparency isn't optional—it's the foundation of trust. Today, Anthropic's revenue figures come from anonymous sources, not audited financials. The company filed a confidential prospectus with the SEC in June, but we don't know the details. The same opacity plagued OpenAI, whose $40 billion run rate also came from unnamed insiders. Neither company confirms how they calculate the metric. In crypto, we call that a lack of proof-of-reserves. Now, the contrarian angle. Revenue growth at this scale is impressive, but it doesn't measure the health of the ecosystem. Anthropic and OpenAI are building walled gardens—closed models, proprietary data, and centralized control. Their financial success is a testament to the market's appetite for AI, but it's also a warning. When a single entity controls a $2 trillion valuation, it becomes a gatekeeper. The very principles of decentralization that underpin blockchain—open protocols, community governance, user sovereignty—are antithetical to this model. In 2020, I organized DeFi workshops to translate complex smart contract mechanics into accessible benefits. The goal was to empower users, not to enrich a centralized team. Today's AI giants are doing the opposite: they're enriching themselves while users hand over data and agency. We didn't ask for a $2 trillion gatekeeper. We asked for open, accountable AI. The 2024 ETF educational initiative I authored highlighted the tension between institutional adoption and core decentralization values. The same tension exists here. Anthropic's IPO will likely be a windfall for early investors and employees, but what about the broader community? The company's Claude models are powerful, but they're not open source. The training data, the weights, the inference infrastructure—all proprietary. Compare that to the Ethereum community, where the protocol is open, upgrades are debated publicly, and anyone can run a node. The AI industry is replicating the same centralized power structures that blockchain was supposed to overthrow. My experience in the 2026 AI-crypto convergence forum reinforced this. We brought together 50 experts to define ethical standards for autonomous economic agents—AI agents interacting with blockchain wallets. The key takeaway was the need for "Human-in-the-Loop" protocols to preserve accountability. But who is the human in the loop for Anthropic? Its board? Its investors? The SEC? We need to demand that these companies adopt transparent governance, open-source contributions, and community oversight. Otherwise, we're just trading one set of centralized gatekeepers for another. Let's look at the numbers more critically. Anthropic's $10 billion in total revenue for 2025 is impressive, but the quarterly revenue more than doubled from $4.73 billion in Q1 to $11.5 billion in Q2. That's a 143% quarter-over-quarter leap. Such growth rates are unsustainable, and when they slow, the stock price could crater. The $2 trillion valuation implies a price-to-sales ratio of roughly 30 times run-rate revenue. That's optimistic, even for a high-growth AI company. The crypto market saw similar froth during the 2021 NFT boom—projects valued at billions based on a few weeks of trading volume. We all know how that ended. We didn't fall for that in 2021, and we shouldn't fall for it now. The same skepticism that drives us to audit smart contracts should drive us to scrutinize AI company financials. The run rate metric is exactly the kind of marketing tool that ICOs used to attract liquidity. Remember the days when projects promised "annualized yields" of 1000% based on a single day of token emissions? That's what run rate feels like: a number designed to impress, not to inform. So where does this leave us? As AI and blockchain converge, we have a choice. We can cheer for Anthropic's market dominance and hope that its ethical alignment research translates into good behavior. Or we can demand that the next generation of AI infrastructure be built on open, decentralized principles. The 2022 bear market taught me that resilience is a communal effort. When the crash came, I partnered with open-source foundations to provide mental health resources and career transition advice. We didn't abandon the community; we strengthened it. The same approach is needed now: build systems that are robust, transparent, and accountable. The takeaway is not to dismiss Anthropic's achievement. It's a remarkable business story. But the real test is what happens after the IPO. Will the company use its capital to double down on closed models, or will it open up its research and governance? Will it become a force for decentralization, or will it become another monopoly? The answer will determine whether the AI-crypto convergence serves humanity or just a handful of shareholders. We didn't build this industry to create new billionaires. We built it to distribute power. Let's not forget that as the next bull run approaches.