Pulse checks from the blockchain veins — but this time, the signal isn’t on-chain. It’s in the boardroom.
Over the past 48 hours, the crypto-native AI narrative has been jolted by a governance play rarely seen in the industry: Anthropic, the $60B+ AI lab, is reportedly planning to issue super-voting shares to its CEO and co-founders ahead of a potential IPO. The move, first reported by Crypto Briefing, is a classic dual-class stock structure — but in the context of AI safety and decentralized alignment, it carries weight far beyond a typical Silicon Valley power play.
Why now?
Anthropic is at a critical juncture. The company has raised over $10B from Amazon, Google, and others, but its valuation growth has outpaced its ability to generate revenue. An IPO is the logical next step to provide liquidity for early investors and employees. But the founders — Dario and Daniela Amodei — are not content to let the market dictate their pace. They want control. And they’re willing to take a valuation haircut to get it.
The super-voting structure is a tool of last resort for founders who fear dilution. In tech history, it’s been used by Google (10-to-1), Facebook (10-to-1), and Snap (no voting rights for public shareholders). But Anthropic’s case is different: the company’s entire brand rests on a promise of responsible AI development. If the founders concentrate power, who holds them accountable?
Core: The Math of Control
Let’s quantify the risk. A typical dual-class structure gives founders 10 votes per share, while public shareholders get 1 vote. If the Amodeis hold 10% of the equity, they could control over 50% of the voting power. That’s a level of concentrated decision-making that should raise red flags for any institutional investor.
Using historical data from academic studies (e.g., Gompers, Ishii, Metrick, 2010), dual-class firms trade at a discount of 5–10% compared to single-class peers. For Anthropic, assuming a pre-IPO valuation of $60B, that discount translates to $3B–$6B in lost market capitalization. That’s a billion-dollar tax on governance risk.
But the real cost is not just financial. It’s strategic. Anthropic’s largest investors — Amazon and Google — are also its competitors in the AI cloud market. If the founders can make unilateral decisions on compute procurement, model release schedules, or licensing terms, the strategic alignment with these hyperscalers becomes fragile. From my own surveillance work during the 2025 AI compute crush, I saw how GPU allocation decisions could make or break a project. Concentrated voting power can accelerate decisions, but it can also amplify errors.
Contrarian: The Safety Paradox
The mainstream narrative is that super-voting shares protect Anthropic’s long-term mission from short-term profit pressure. The founders are known for their commitment to AI safety — they left OpenAI over concerns about its commercialization. But this is exactly where the paradox lies.
Tracing the ICO gold rush scars — I remember 2017, when ICOs promised decentralized governance but delivered founder-controlled wallets. The same pattern repeats: a mission-driven founding team claims to need super-voting rights to protect the vision, but the result is a lack of external accountability. In AI safety, where the stakes are existential, the absence of independent oversight could be catastrophic.
Consider the Terra/Luna collapse in 2022. I was monitoring whale wallets when the depegging hit. The core issue was not a technical flaw but a governance failure: Do Kwon had unilateral control over the protocol’s mechanics. Concentrated power in a system claiming to be decentralized is a ticking time bomb. Anthropic’s super-voting structure is a milder version of the same logic — a promise of safety paired with a mechanism that reduces safety checks.
Moreover, the super-voting plan may alienate the very investors who care most about ESG and governance. BlackRock, Vanguard, and State Street have all publicly stated they will vote against dual-class structures in new listings. If Anthropic needs their capital to support a $60B+ float, the governance discount could widen to 15% or more.
Takeaway: The Sunset Clause is the Key
Arbitrage angles in chaotic markets — The smart money is not betting on the IPO price; it’s betting on the terms. The super-voting plan is almost certainly accompanied by a sunset provision: a clause that terminates the extra voting power after a certain period or upon the founders’ departure. The question is: what is the sunset? If it’s 10 years, that’s a long time for a young AI company. If it’s tied to specific safety milestones, that could be a first-of-its-kind governance innovation.
Anthropic is not just another tech IPO. It’s a test case for how AI companies balance control with accountability. If the founders succeed in maintaining super-voting rights without a meaningful sunset, they will have created a new class of AI aristocracy. If they fail, the market will impose its own discipline.
Cheetah pace against systemic collapse — The next signal to watch is the S-1 filing. Look for the exact voting ratio, the sunset clause language, and any independent board oversight. Until then, the super-voting plan is a governance black hole: we know it’s there, but we can’t see its boundaries.
For crypto-native investors, this is a reminder that the most important decentralization is not just in code — it’s in power. And power, left unchecked, eventually collapses.