Anthropic Founders Demand 50.1% Voting Control Ahead of IPO
TL;DR
- Anthropic is asking existing shareholders to approve a plan giving its seven co-founders 50.1% voting control through a new class of supervoting shares, locking in founder control ahead of a possible IPO.
- The company says the move is needed to protect its long-term AI safety mission from short-term public market pressure and hostile takeovers, echoing the dual-class structures used by Meta, Google, and Snap.
- Investors and governance experts are split, with some backing mission-driven independence while others warn the unprecedented concentration of power could weaken accountability and set a risky precedent for public shareholders.
The Power Play
Anthropic has begun a high-stakes campaign to reshape its own governance before it ever rings a public-market bell.
According to reports this week, the AI startup is asking current shareholders to sign off on a restructuring that would issue a special class of shares granting its seven co-founders — including CEO Dario Amodei, President Daniela Amodei, Jack Clark, Sam McCandlish, Tom Brown, Jared Kaplan, and Ben Mann — a combined 50.1% of voting power.
In practice, that means majority control over board elections, mergers, acquisitions, and any future charter changes, even after an initial public offering would massively dilute their economic ownership. The proposal requires shareholder approval in the coming weeks, setting up a critical test of investor confidence in one of the most valuable private companies in the world.
The move comes as Anthropic lays the groundwork for what could be a blockbuster IPO as early as 2026 or 2027. The company was last valued at around $183 billion in its late-2025 private raise and has been in talks for additional funding at valuations reportedly topping $300 billion, fueled by explosive growth of its Claude models in enterprise and coding markets.
Why Anthropic Wants Founder Control Now
Anthropic leadership has framed the request not as a power grab, but as mission protection.
Since its 2021 founding as a public benefit corporation, Anthropic has positioned itself as the safety-first alternative to OpenAI, complete with an unusual governance layer that includes its Long-Term Benefit Trust, which can appoint independent directors, and published Responsible Scaling Policy for frontier models.
Sources close to the company say founders fear that once public, Anthropic could face intense pressure from activist investors, short-term earnings demands, or even a hostile takeover bid from Big Tech rivals or sovereign-backed funds — all of which could force it to cut corners on safety to chase growth.
By securing 50.1% voting control now, while still private, the founders argue they can guarantee the company stays focused on building safe artificial general intelligence over decades, not quarters. The model is familiar in Silicon Valley: Mark Zuckerberg controls Meta with around 61% of voting power, Larry Page and Sergey Brin retain majority control at Alphabet, and Evan Spiegel and Bobby Murphy control Snap.
Anthropic executives have also pointed to the chaos of OpenAI’s 2023 boardroom ouster of Sam Altman as a cautionary tale of what happens when AI governance breaks down under pressure.
How Investors Are Reacting
Early signals suggest most major backers are leaning toward approval, but not without grumbling.
Anthropic’s cap table includes Amazon, Google, Menlo Ventures, Lightspeed, Spark Capital, and sovereign fund MGX, many of which have poured billions into the company and stand to gain enormously from an IPO. For them, founder control is the price of admission to the AI boom.
Several private investors have privately expressed support, arguing that founder-led companies outperform over the long run and that Anthropic needs stability to compete with OpenAI, xAI, and Google DeepMind. One investor described it as a “take it or leave it” moment — rejecting the plan could delay or derail an IPO.
Others, however, are wary of signing away rights with limited financial disclosures and no public vote. Some late-stage funds worry about downside protection: if Claude’s growth stalls or safety incidents emerge, they would have almost no mechanism to force leadership changes.
What Governance Experts Are Saying
Corporate governance watchdogs are sounding louder alarms.
Critics argue Anthropic’s proposal goes further than even Meta or Alphabet, concentrating absolute majority control in the hands of seven people rather than one or two founders, with no sunset clause reported so far. Sunset provisions, which phase out supervoting shares after 7-10 years or when founders leave, have become best practice for modern tech IPOs to balance vision with accountability.
Experts also note the timing is unusual. Most startups impose dual-class structures at the IPO filing stage, not via a retroactive private shareholder vote. Asking private shareholders to approve it early could limit the leverage public-market investors would normally have through S&P and Nasdaq stewardship guidelines, proxy advisors like ISS and Glass Lewis, and large index funds that routinely vote against perpetual founder control.
Supporters counter that AI labs are a special case. They argue traditional shareholder democracy is ill-suited to companies building potentially transformative technology, where a single rushed deployment decision could have global consequences.
What It Means for a Future IPO
If approved, the 50.1% structure would fundamentally shape Anthropic’s public market debut.
On the upside, it would give Anthropic a clear, stable story for its roadshow: invest in our growth, but don’t expect to steer the mission. That clarity helped Meta and Palantir attract long-term holders despite governance concerns, and could allow Anthropic to command a premium valuation by promising immunity from activism.
On the downside, it could narrow the pool of buyers. Many ESG-focused funds, pension funds, and index providers penalize or exclude companies with unequal voting rights. Institutional investors may demand a discount, stronger independent board commitments, or binding safety audits as a tradeoff for ceding control.
The proposal also raises unresolved questions: How long will the supervoting shares last? What happens if a co-founder departs? Will the Long-Term Benefit Trust retain its powers or be sidelined? And how will public shareholders enforce Anthropic’s public benefit charter if founders hold an unbreakable majority?
The Road Ahead
Anthropic has not publicly confirmed a timeline for a vote or an IPO filing, and the company declined to comment in detail on the reports.
What is clear is that the next few weeks will be decisive. If shareholders approve, Anthropic will head toward Wall Street with one of the most founder-friendly governance structures in tech history — a bet that mission lock-in matters more than shareholder democracy.
If they push back, it could force a renegotiation that delays one of the most anticipated IPOs of the AI era.
Either way, Anthropic’s bid for 50.1% control is more than a corporate technicality. It is a referendum on who should control the companies building superintelligent AI: the founders who started them, the investors who fund them, or the public who will ultimately live with the consequences.
Get All The Latest Updates Delivered Straight To Your Inbox For Free!