CIA's In-Q-Tel bought startup equity instead of writing government contracts

By the late 1990s commercial markets, not government labs, set the pace of information technology, and the CIA's procurement could not compete at internet speed. Ruth David and Joanne Isham saw the gap, and DCI George Tenet gave the idea cover; in 1998 a small team took the concept to Silicon Valley to be put through the wringer. Norman Augustine chaired the new corporation: legally formed in February 1999, first Agency contract in March, unveiled on September 29, 1999 as In-Q-It — renamed In-Q-Tel that January — with Gilman Louie as CEO and a first-year budget near $28 million.

The design was a bundle of deliberate contradictions, as a 2005 Washington Post profile put it: independent of the CIA yet answerable to it, a nonprofit whose people can profit, public in identity while secret in product. The mechanics came from venture capital: In-Q-Tel could take equity, form joint ventures, fund grants, sponsor competitions and award sole-source contracts, without case-by-case Agency approval. It invested only where an Agency need overlapped real commercial demand, and only in unclassified work at first — so startups would take the call.

Results compounded fast. Within five years In-Q-Tel had invested in more than 75 companies and delivered over 100 technologies to the CIA on roughly $37 million a year; one investee called it a buffer from being a real government contractor. Its most famous win was Keyhole, the satellite-mapping startup Google bought in October 2004, whose technology became Google Earth; In-Q-Tel's stake later cashed out at over $2.2 million in Google shares. By 2016 it had made about 325 investments on at least $120 million a year, and the Army and NASA were building copycats.

Buy startups like a VC, not like a contractor

The instrument matched the market: equity and market-rate deals let the agency transact at startup speed instead of procurement speed, because In-Q-Tel negotiated without case-by-case Agency approval.

Incentives aligned both ways — startups kept their commercial upside and only worked on unclassified problems, so Silicon Valley would take the call.

Investing at the intersection of Agency need and commercial interest leveraged private capital: the CIA paid a slice and got technology the whole economy was funding anyway.

Hard Agency problems doubled as product roadmaps for investees, turning the customer's pain into the vendor's market advantage.

When a buyer is slower than its market, don't fix the paperwork — change the instrument. Paying on the market's terms motivates unwilling vendors, and private capital leverages public money.

The experiment became an institution: about 325 investments by 2016 per the Wall Street Journal, an expanding budget, and a model copied by the Army, NASA and other defense and intelligence agencies within six years of launch. The portfolio seeded some of the industry's defining tools — the Keyhole investment that became Google Earth is the canonical example — and the nonprofit later held roughly a billion dollars in assets on its 25th anniversary, still operating on the founding bargain: venture terms, unclassified work, commercial upside, agency outcomes.