San Francisco real estate transactions are becoming venues for speculative bets on private artificial intelligence companies. Home sellers in the Bay Area now demand equity stakes in OpenAI and Anthropic as partial payment for properties, bypassing traditional cash offers entirely.
This dynamic reflects the extraordinary private valuations commanding the AI sector. OpenAI's most recent funding round valued the company at $80 billion. Anthropic, its competitor, raised capital at a $15 billion valuation. Neither company has gone public, yet their equity has become liquid enough in secondary markets that homeowners treat it as acceptable tender for multimillion-dollar purchases.
The practice signals how concentrated wealth has become among AI industry insiders and early investors. Employees at these companies hold restricted stock units and options that have appreciated sharply without a public exit event. For sellers, accepting equity offers a potential upside if these companies eventually complete IPOs. For buyers, it provides a mechanism to deploy concentrated equity holdings without triggering immediate tax consequences on the full sale price.
Real estate agents report that this trend accelerates competition for high-end properties in San Francisco and nearby areas like Palo Alto and Mountain View. Sellers can demand equity precisely because multiple bidders hold substantial positions in the same private companies. A buyer with $5 million in unvested OpenAI options can structure a purchase as partial equity transfer rather than liquidating assets.
The practice carries substantial risk. Private equity in unproven companies, however well-capitalized, remains illiquid and speculative. A buyer betting their home purchase on an Anthropic IPO faces execution risk if the company delays going public or underperforms expectations. Sellers accepting equity gamble that their equity stake will retain value and that they can eventually exit it.
This phenomenon also highlights regulatory gaps. Secondary markets for private equity exist in fragmented, opaque channels. Unlike public stock exchanges, there is minimal transparency around valuations or transaction volumes in companies like OpenAI and Anthropic. A buyer and seller agreeing on a stock price for a home may not reflect actual market consensus.
The trend will persist as long as AI companies remain private with high valuations. If OpenAI and Anthropic delay IPOs, expect more Bay Area sellers to demand equity. If either company goes public at steep premiums, early sellers will claim prescient timing. Either way, San Francisco's real estate market now trades on the fortunes of private AI firms.