The $45 Billion Decision That Shaped Who Owns AI and Space Infrastructure
In 2015, Vishal Sikka, then Infosys's chief executive, tried to convince the company's board to invest $1 billion in a small San Francisco research lab called OpenAI. The board declined, according to reporting by the Times of India. Ten years later, that decision may be the most expensive corporate governance call in the history of Indian technology.
OpenAI, now preparing a public listing that could value it at $850 billion or more, filed its S-1 with the SEC this week, CNBC confirmed. The Times of India calculates that if Sikka's proposal had proceeded, a $1 billion investment at OpenAI's 2015 valuation could be worth roughly $45 billion today — nearly 90 percent of Infosys's entire market capitalization. That calculation uses OpenAI's 2015 valuation as the entry-point proxy; neither OpenAI nor Infosys has publicly confirmed the number, and the estimate has not been independently verified. The magnitude is not in dispute; the precision is.
Two Indian-American investors made the opposite bet. Vinod Khosla, through Khosla Ventures, placed $50 million into OpenAI in 2019 as one of the earliest major backers of the company behind ChatGPT. That stake could now be worth roughly $1.5 billion, a 30-fold return, the Times of India reported. Anand Desai founded Darsana Capital Partners in 2014 with a stated focus on long-duration technology bets, positioning the fund toward companies with multi-year commercialization horizons rather than near-term exits. He entered SpaceX at a $30 billion valuation in 2019; Darsana has roughly 60 percent of its $15 billion in assets under management invested in the rocket company, and its paper gains could top $10 billion at SpaceX's target valuation of up to $1.75 trillion, the Wall Street Journal reported.
SpaceX is also preparing to go public. Its S-1 filing is on file with the SEC.
The coincidence is remarkable: two diaspora investors positioned at the center of the two most consequential infrastructure platforms of the next decade, both cashing out in the same season. But the diaspora framing obscures the more immediate point. Both Khosla and Desai made their bets during a period when the Indian technology establishment was looking elsewhere — toward consumer internet, outsourcing contracts, and domestic market plays. The investors who placed early capital in frontier AI and space infrastructure were not the consensus picks of their own professional community.
What the IPOs will not change is who holds the keys. A startup building on OpenAI's API stack pays prices OpenAI sets. A satellite operator buying launch capacity from SpaceX joins a queue SpaceX manages. When infrastructure platforms concentrate ownership among long-duration investors, the incentives shaping platform governance differ from those of diversified public shareholders. Whether that difference translates into concrete access or pricing effects is an open question neither company has addressed publicly. Neither Khosla's firm nor Darsana responded to requests for comment on their post-IPO intentions.
The Sikka board decision, regardless of whether the $45 billion estimate is accurate, is a proxy for a larger question Indian technology never answered: whether to treat frontier AI and space infrastructure as national strategic assets or as ordinary venture bets. Indian tech's established firms chose the latter. Diaspora investors chose the former. The consequences of that split are about to become legible in the shareholder registers of the two most important technology platforms of the next decade. When diaspora investors who made early, high-conviction bets become among the largest outside shareholders of AI and space infrastructure, every builder who pays OpenAI's API prices or buys SpaceX launch capacity is, in a small way, funding that network's returns. Whether that concentration of ownership shapes platform governance in practice is a question the IPO disclosures have not yet answered.