In early 2025, Mira Murati, the former Chief Technology Officer of OpenAI, quietly launched what may become one of the most influential AI companies of the decade: Thinking Machines Lab. With little more than a website and a roster of elite talent, the company has already attracted the attention—and capital—of Silicon Valley’s most powerful players. In a deal led by Andreessen Horowitz (a16z), Thinking Machines is reportedly raising $2 billion at a staggering $10 billion pre-product valuation.

What’s even more striking than the numbers is the deal structure. According to The Information, Murati is not only seeking a minimum investment of $50 million per backer—a high bar even by AI standards—but is also retaining an extraordinary level of control over the company. In addition to typical founder-friendly terms like dual-class shares, Murati is negotiating “superpower rights” that would allow her to make strategic decisions unilaterally, override board input, and shape the company’s trajectory with minimal external interference.

What Murati Negotiated—and What It Means

According to The Information, Mira Murati has secured one of the most founder-favorable deal structures seen in recent AI fundraising. Among the key elements she pushed through:

• Supervoting shares that grant her outsized voting power, even as large institutional investors come on board. This ensures that she can retain control over major company decisions—including hiring, fundraising, and product direction—regardless of dilution.
• Board control provisions that may give her the ability to appoint or remove board members, or to prevent changes in board composition without her consent. This effectively neutralizes investor influence at the governance level.
• No performance-based vesting tied to her equity. Unlike typical founder agreements that tie ownership to time or milestones, Murati reportedly retains her equity outright, offering her maximum flexibility and leverage.
• Minimal investor rights in areas such as information access, liquidation preferences, or veto powers. While the exact terms remain private, sources suggest that even lead investors like Andreessen Horowitz accepted weaker-than-usual protective provisions.
• A high minimum investment threshold—reportedly $50 million per investor—which not only filters for deep-pocketed backers but also reduces cap table complexity and potential opposition.

Collectively, these terms point to a founder-first model with almost no formal constraints. In Silicon Valley, such terms are typically reserved for founders with proven ability to deliver transformational outcomes. Murati appears to have joined that elite circle—one where capital follows conviction, and governance yields to vision.

The excitement surrounding the deal is emblematic of a broader dynamic in tech: the growing concentration of capital around a small number of “founder kings” (or queens) who command not only attention but also structural dominance over their companies. With AI being the hottest sector in tech, and Murati carrying the aura of having shaped GPT-4 at OpenAI, investors are lining up—despite the loss of control that comes with the ticket.

Why Investors Are Saying Yes

Murati’s credentials are hard to match. As OpenAI’s former CTO, she played a central role in shaping some of the most advanced large language models to date. Her new venture, Thinking Machines Lab, promises to build AI systems that are more customizable, broadly capable, and accessible—a subtle but intentional contrast to the tightly controlled and opaque nature of current models like ChatGPT and Google Gemini.

Her founding team is a who’s who of OpenAI veterans, including Bob McGrew (former research director) and Alec Radford (co-creator of GPT-2 and GPT-3). For many investors, the logic is simple: if OpenAI was a $100 billion opportunity, then a spin-out led by its former brain trust could easily be worth $10 billion or more—especially in a market eager for alternatives.

The Investor Dilemma: Power vs. Protection

But beneath the enthusiasm lies a thorny governance debate. Murati’s deal terms introduce “superpower rights” that go beyond Silicon Valley’s usual founder-friendly frameworks. While such arrangements are intended to protect visionary leadership, they introduce a range of risks and trade-offs for investors.

On the positive side, these rights allow Murati to shield the company from short-term investor pressure, enabling long-term bets that may take years to materialize. In a field as volatile and technically complex as AI, such autonomy can be essential. Some investors even argue that micromanagement by non-technical board members is a bigger risk than founder dominance, particularly at the earliest stages of radical innovation.

However, the downsides are significant—and often underappreciated:

1. Loss of governance leverage
   Investors may find themselves effectively unable to intervene if the company missteps—be it in product strategy, financial management, or compliance. With superpower rights in place, a misaligned or overly aggressive strategy might go unchecked until it’s too late.
2. Reduced transparency
   When founders control not just the votes but also the information flow, investors are left with limited visibility into key developments. This asymmetry can make active portfolio management difficult, especially for institutional funds with fiduciary duties.
3. Future fundraising friction
   Later-stage investors often demand clearer governance terms and protection clauses. Superpower rights at the seed or Series A level may deter follow-on capital from more risk-averse investors—particularly public market crossover funds or sovereign wealth vehicles.
4. Team and culture risks
   A power structure that concentrates authority exclusively at the top can alienate talent, especially senior hires who expect a seat at the table. Over time, this may hinder innovation or lead to attrition among key contributors.
5. Exit complexity
   In an acquisition scenario, acquirers typically require clean governance structures and board-level alignment. A founder with unilateral control may complicate negotiations or even block a lucrative exit, putting investor returns at risk.

Ultimately, the structure places a huge amount of trust in a single individual. While Murati’s track record arguably justifies that trust, the precedent being set is clear: capital will flow where talent goes—even if it means surrendering control.

A New Template for Elite Founders?

The Thinking Machines deal may be remembered not just for its eye-popping valuation, but for codifying a new norm in frontier tech investing: that some founders now have the leverage to write their own rules. Whether this results in legendary returns—or a cautionary tale—will depend not only on the science, but on Murati’s judgment, adaptability, and leadership.

For investors, the calculus is simple—but unforgiving: back the outlier, accept the risk, and hope you’re part of history in the making.