Since 1999, a nonprofit called In-Q-Tel has seeded over 300 tech companies. The products they created were never only what they appeared to be.

On the 29th of September, 1999, a non-profit corporation was registered in the state of Virginia. The paperwork was ordinary: a charter, a board structure, an operating mandate. Four pages. The language was careful and familiar, the kind of language that announces nothing. It authorized the corporation to receive funds from the United States government. To invest those funds in private technology companies. To return any proceeds not to the Treasury but to a successor fund that would invest again.

The charter did not use the word “surveillance.” It used the word “innovation.”

The distinction between those two words, sustained over the next twenty-five years, would become one of the most consequential ambiguities in the history of American technology policy. Not because anyone was deceived. Every transaction In-Q-Tel ever made is, in some form, a matter of public record. But because the language of investment, of portfolio, of commercial partnership, performed a specific function: it transformed what was a procurement decision into what looked like market behavior. And market behavior, in the grammar of modern power, has no authors.

The corporation’s name was In-Q-Tel. Its sole customer was the Central Intelligence Agency.

The architecture

By the late 1990s, the CIA’s technology development system was running fifteen to twenty years behind the commercial sector. Its ability to process open-source information, run pattern analysis across large datasets, or parse multilingual documents at scale was falling behind. In 1998, CIA Director George Tenet commissioned a report from the Business Executives for National Security. The conclusion: the CIA needed to behave like a venture investor. Fund the companies engineers were already building, with terms that required the companies to share what they built.

They chose Gilman Louie as the first CEO. A video game executive. No intelligence background. That was, in its own way, the point. The fund would look like Silicon Valley because it would recruit from Silicon Valley, behave like Silicon Valley, and speak Silicon Valley’s language.

Its legal structure as a 501(c)(3) nonprofit is not a technical detail. It is the structural keystone of the entire operation. A nonprofit whose revenues come from a classified government budget and whose investments are disclosed selectively occupies a category that existing oversight frameworks were not designed to scrutinize. Not secret. Not private. Something between.

The oversight gap is not a hole in a fence. It is a property of the architecture.

The investments

In 2003, a company in Mountain View called Keyhole Corporation was running out of money. It had built EarthViewer: software that allowed a user to navigate a three-dimensional model of the entire surface of the earth from a laptop. A call came from Virginia. Keyhole accepted the investment.

Within months, Keyhole’s technology was in operational use by American military and intelligence planners. In April 2003, during the invasion of Iraq, military units were using EarthViewer to navigate urban terrain in Baghdad.

In October 2004, Google acquired Keyhole. In June 2005, Google released the product to the public under a new name: Google Earth.

The word that describes this transition in public discourse is “commercialization.” The word implies a neutral sequence: innovation, followed by adoption, followed by diffusion. What the Keyhole sequence shows is a different order entirely: procurement, followed by development, followed by commercial release. The government did not adopt a technology. It seeded one, classified it, tested it, and then released it into civilian life through a private acquisition.

In 2004, Palantir Technologies received an early In-Q-Tel investment. Two million dollars. Within the markets that connect Silicon Valley to the national security state, that amount functioned as a credential. Not a security clearance in the legal sense, but a certification: the CIA’s venture arm had evaluated this company and found it suitable for work in sensitive environments. For government procurement officers, it reduced hesitation. For the company itself, it opened doors that remained closed to competitors without the signal.

By the time Palantir listed on the New York Stock Exchange in September 2020 at a valuation peaking above fifteen billion dollars, sixty-seven percent of its revenue came from government clients.

The two million dollars did not build Palantir. It determined what Palantir was allowed to become.

The diffusion

By 2016, In-Q-Tel had made cumulative investments in more than three hundred companies. The portfolio touches nearly every layer of the modern technology stack: machine learning, computer vision, natural language processing, biometrics, geospatial analysis, network analysis, cloud infrastructure, and satellite imagery.

In 2016, a company called Geofeedia was providing social media surveillance tools to law enforcement agencies across the United States, allowing police to monitor public social media posts by geographic location. Departments in Baltimore, Oakland, Chicago, and elsewhere used it to monitor Black Lives Matter demonstrations. The ACLU documented this. Facebook, Twitter, and Instagram revoked Geofeedia’s data access. Geofeedia was dissolved in 2017.

Geofeedia was an In-Q-Tel portfolio company.

No one at In-Q-Tel directed American police to use Geofeedia against protesters. The mechanism was producing its outputs exactly as designed. Each step in the chain was legal. Each participant behaved according to its incentives. No one was coordinating.

That is the point.

The full investigation traces this further.

What follows changes the picture in a specific way. The full piece maps the closed loop of public capital, private returns, and self-sustaining reinvestment. It addresses the strongest counter-argument: whether the causation might run in the opposite direction. And it traces what In-Q-Tel did not fund, and why those absences matter as much as the investments.

The complete investigation is published at The Manifest Archive.

Read the full article here or subscribe for free to receive every piece directly in your inbox.