Anthropic's IPO Push: Founder Control, $11.6B Deal, Biology

A dense 48 hours for Anthropic watchers: the company floated a Palantir-style dual-class share structure, signed one of the largest compute contracts in AI history, and disclosed a significant result from its internal biology lab, all as its IPO preparations accelerate.

Key points

• Anthropic proposes special-class shares that would give its seven co-founders collective voting control of 50.1%, regardless of their economic ownership at listing.
• The company has signed a seven-year, $11.6 billion computing agreement with Akamai to secure large-scale infrastructure capacity as it scales model development.
• Anthropic’s biology lab has already produced a major discovery, with Claude kept under human supervision throughout the research process.
• The coming IPO is mobilizing competing political donor networks, with wealthy backers and opponents clashing over AI regulation ahead of the listing.
• Early investor Bessemer has publicly praised Anthropic’s consistent safety posture, flagging it as a key reason behind $5.75 billion in growth-stage AI commitments.

What does 50.1% founder control mean for public investors?

The proposed voting structure, reported by The Information, would concentrate decisive governance authority in the hands of the seven co-founders as a bloc, even if their combined equity stake shrinks substantially through dilution and secondary sales over time. The Palantir comparison is instructive: that company used a similar mechanism to let its founders retain strategic control well after listing, which pleased some long-term investors and frustrated others who wanted more board accountability.

For Anthropic, the structure carries a particular wrinkle. The company’s core identity is built around safety-conscious development and long-termism. Founders retaining voting control could be framed as a guardrail against short-term shareholder pressure to cut safety spending or accelerate deployment. Public market investors, though, will need to weigh that framing against the reduced leverage they would have over management decisions. Institutional buyers who typically push for one-share-one-vote standards may negotiate hard during the roadshow.

The Akamai deal: what $11.6 billion in compute signals

The seven-year agreement with Akamai is a striking commitment on both sides. At roughly $1.66 billion per year in annualized compute spend, it sits among the largest infrastructure contracts any AI lab has signed to date. For Anthropic, locking in capacity at scale ahead of an IPO serves two purposes: it reduces the risk of being outbid for GPU and data-center access during a crunch, and it gives prospective public shareholders a clearer picture of the company’s cost structure and infrastructure runway.

For Akamai, the deal represents a significant pivot deeper into AI-grade compute, a market the company has been building toward through prior data-center acquisitions. The seven-year term means both parties are betting on sustained demand for frontier model training and inference at least through the early 2030s. Anthropic has not disclosed the pricing mechanics or whether the contract includes capacity guarantees, so investors should watch for those details in any forthcoming S-1 filing.

A biology lab result and a political money fight: two IPO-era subplots

TechCrunch reports that Anthropic’s biology lab has surfaced a significant scientific result, with Claude operating under continuous human oversight. Anthropic has not disclosed what the finding is, but the announcement is clearly timed to demonstrate that its AI-assisted research model can produce real scientific value while maintaining the human-in-the-loop protocols the company has publicly championed. For investors, it is an early signal of the life-sciences revenue opportunity that could diversify the company beyond enterprise software and API access.

On the political side, Bloomberg details how the anticipated IPO wealth is already reshaping AI policy advocacy. A $140 million pro-AI PAC haul has drawn a countermobilization from donors pushing for stricter safeguards, creating pressure on Congress and regulators at precisely the moment Anthropic would prefer a predictable legislative environment. Regulatory uncertainty has historically widened IPO risk premiums in technology listings. How much of that uncertainty crystallizes into actual policy before Anthropic prices its shares remains an open question, and one that underwriters will be watching closely.

Bessemer’s public comments add a layer of investor sentiment data. The firm told Bloomberg that Anthropic’s steady safety stance is a feature, not a cost center, citing it as a rationale for $5.75 billion in broader growth-stage AI bets. That framing aligns with how Anthropic itself is likely to pitch the public market: safety as competitive differentiation, not regulatory burden.