In-Q-Tel: How the CIA Built a Venture Fund, and Where t...

When people picture the way public money reaches a technology company, they usually picture a contract: an agency writes a solicitation, vendors bid, one wins, and the paperwork is a matter of record. In-Q-Tel does something different. It takes intelligence-community money, buys small equity stakes in private startups, and sits on the same cap table as commercial venture capitalists. The mechanism is unusual enough, and the portfolio is classified enough, that it is worth walking through slowly: what the entity actually is, how the dollars move, which bets became famous, and what the documented oversight record says on both sides.

This piece keeps documented facts in its own voice, and it keeps contested judgments strictly attributed to whoever asserted them. Where a number is an estimate, it is labelled an estimate. Where a finding comes from a newspaper rather than an audit, it is attributed to the newspaper.

What In-Q-Tel is, and is not

In-Q-Tel is a private, independent 501(c)(3) nonprofit. It is not a government agency, and that distinction carries most of the weight in everything that follows. It was chartered in 1999 at the request of the CIA, under Director of Central Intelligence George Tenet, to close the gap between the intelligence community and the fast-moving commercial technology sector in Silicon Valley. Its first chairman was Norman Augustine, the former Lockheed Martin chief executive, and its first CEO was Gilman Louie, who had come out of the video-game industry.

The name is a joke that stuck. It plays on "Q", the fictional gadget-master who supplies James Bond, according to former CIA deputy director Sue Gordon, who confirmed the reference to Fortune. The venture cycled through earlier working names, Peleus and then In-Q-It, before landing on In-Q-Tel. It is headquartered in the Northern Virginia suburbs of Washington. Wikipedia places it in Tysons; its IRS Form 990 lists McLean. Both are unincorporated parts of Fairfax County, so there is no single street address worth treating as canonical.

The shorthand "the CIA's venture capital arm" is accurate as far as it goes, but it undersells the scope. In-Q-Tel serves the broad intelligence community, which includes agencies such as the DIA, FBI, NGA, and NSA, not the CIA alone.

How the money moves

The money flow has four steps, and each one has a public proof point or a clearly labelled estimate.

Step one: Congress funds the intelligence community, and a slice reaches In-Q-Tel as government grants and contracts. Because In-Q-Tel files public Form 990s as a nonprofit, its total revenue is visible even though the underlying intelligence appropriation is not. The exact totals, confirmed against IRS Form 990 filings via ProPublica under EIN 52-2149962, are $184.1 million for the fiscal year ending March 2023, $108.26 million for the year ending March 2024, and $132.1 million for the year ending March 2025. Those swings are large, and they are driven partly by investment gains and losses and by the timing of grants, so no single year should be read as "the budget." The commonly cited figure of roughly $100 million a year in government funding is an approximation, not an appropriated line item. The government-only share within each year's revenue is reported but was not itemised in the summary used here, so the totals are the firm ground and the government-only sub-figures should be treated more cautiously.

Step two: In-Q-Tel makes small equity and product-development investments in startups. Individual checks run roughly $500,000 to $3 million, and In-Q-Tel almost always co-invests alongside private venture capitalists rather than taking a controlling stake. Fortune reported, in a July 2025 profile, that for every $1 In-Q-Tel puts in, commercial venture firms add about $40, and that the firm screens on the order of a thousand startups a year and has backed more than 800 companies over its history. Those specific ratios and counts are Fortune's reporting, not audited figures. Fortune also estimated that In-Q-Tel has invested at least $1.8 billion since 1999, describing it explicitly as an estimate drawn from 25 years of tax disclosures. In-Q-Tel declined to confirm the number.

Step three: the startups adapt their commercial technology for government use. In-Q-Tel's stated measure of success is not profit but adoption, meaning pilots and deployments by intelligence-community users. As CEO Steve Bowsher put it to Fortune, "If we make money as part of that, that's great, but that's not the primary goal." (That quote is attributed to Fortune's reporting.)

Step four: when a bet pays off, the proceeds cycle back into the mission. In-Q-Tel does realise gains, and it does not distribute them to shareholders because there are none. It reinvests them. The clearest documented example of a realised gain is the Google share sale described below.

That reinvestment feature is exactly where the honest critique and the honest defence begin to diverge, and it is worth holding the question in mind: when public money buys a stake that later multiplies, who shares in the upside?

The notable bets

Two investments account for most of In-Q-Tel's public reputation, and both are easy to describe imprecisely, so precision matters.

Keyhole, and the road to Google Earth

In 2003, In-Q-Tel invested in a satellite-mapping startup called Keyhole Inc. Google acquired Keyhole in 2004 and relaunched its software as Google Earth in 2005. The precise version of the story is this: In-Q-Tel backed Keyhole, the company. It never owned Google Earth, the product. When Google bought Keyhole, In-Q-Tel's stake converted into Google shares, and Wikipedia reports that In-Q-Tel later sold roughly 5,636 of those shares for about $2.2 million in November 2005.

The size of the original investment is often given as about $527,000, a figure that traces to secondary reporting, including a widely cited account by mapping-industry writer Joe Morrison. One framing of that number describes $527,500 raised across early rounds that included In-Q-Tel and Sony, which means the half-million-dollar figure may not have been In-Q-Tel's check alone. The safe description is a roughly half-million-dollar strategic investment that returned several times its value, with the specific return figures attributed to secondary reporting. Before it was a consumer product, the Keyhole technology served the Pentagon's National Imagery and Mapping Agency and aided U.S. troops in Iraq, which is the mission logic In-Q-Tel points to.

Palantir

Palantir Technologies received early In-Q-Tel funding around 2004 to 2005, an investment of roughly $2 million. The CIA was reportedly Palantir's only customer from roughly 2005 to 2008, a detail corroborated by Forbes, Fortune, and Wikipedia. The proportion matters here: In-Q-Tel was an early and small backer, not Palantir's largest investor. Peter Thiel's own capital dwarfed the In-Q-Tel stake. The dates vary between sources, so "around 2004 to 2005" is the honest range.

Beyond those two, In-Q-Tel's publicly reported portfolio includes names such as Anduril and Databricks, and by one 2016 count the firm listed 325 investments. More than 100 of its investments are classified. Those classified holdings cannot be audited by the public, and no specific classified product or use should be asserted as fact.

The oversight record

Here is where the record has to be read carefully, because a common version of the In-Q-Tel story leans on a document that does not appear to exist. There is no standalone 2016 Government Accountability Office report on In-Q-Tel's finances or conflicts of interest that could be located for this piece. The two solid oversight anchors are a newspaper investigation and an independent panel, and they point in different directions.

The 2016 Wall Street Journal investigation

In August 2016, a Wall Street Journal investigation by reporter Damian Paletta examined In-Q-Tel's governance. It reported that nearly half of In-Q-Tel's trustees had a financial connection to a company that In-Q-Tel had funded, and that trustee-connected firms were funded on 17 occasions. The original article is paywalled, and the finding is corroborated by multiple secondary summaries and by Wikipedia.

The framing has to be exact. These were reported potential conflicts of interest, not adjudicated wrongdoing and not any legal finding. The Journal documented a governance pattern and raised questions about it. It did not establish corruption, and this piece does not characterise it as such. Attribute the finding to the Journal, weigh it as a governance concern, and stop there.

The 2001 BENS panel

Earlier, a congressionally mandated review reached a more favourable conclusion. Under a directive tied to the fiscal year 2000 process, an independent 30-member panel organised by Business Executives for National Security, or BENS, assessed In-Q-Tel and released its report in 2001. The panel judged that In-Q-Tel's potential advantage to the CIA outweighed the risk, a conclusion often summarised as "worth the risk." BENS is a pro-national-security business nonprofit, which is worth naming so readers can weigh the source. The same panel also flagged a real weakness: the CIA lacked a fast internal process to insert the technology it acquired, meaning the front-end investment was outpacing the back-end adoption.

The transparency question

Underneath both of those is a structural point. In-Q-Tel is a private nonprofit. It files public Form 990s, but it is not a government agency, and it is not fully subject to the Freedom of Information Act in the way an agency would be. It discloses only limited information about its investments, and more than 100 of them are classified. The Journal reported that In-Q-Tel sometimes does not reveal an investment at all. So the public can see the top-line revenue and the executive-compensation disclosures, but it cannot audit most of the portfolio, the terms of individual deals, or the full flow of returns.

On compensation specifically, the Form 990 provides its own texture. For the fiscal year ending March 2025, CEO and President Steve Bowsher's reported total compensation was roughly $3.96 million, comprising about $2.44 million in base pay plus roughly $1.52 million in other compensation and benefits. The filings also disclose conflict-of-interest transactions and first-class or charter travel for key employees. Those are the nonprofit's own disclosures, on its own public return.

The same cloth: the other government venture arms

In-Q-Tel worked well enough that the model spread, and the strongest evidence of its perceived success is imitation. But "spread" does not mean the copies are identical. The mechanisms differ, and lumping them together as "government VC" gets the money wrong. Four distinctions are worth keeping straight.

OnPoint Technologies (U.S. Army, 2002 to 2003): equity. OnPoint is the Army's venture-capital arm, authorised under Section 8150 of the fiscal year 2002 Department of Defense Appropriations Act, Public Law 107-117, which earmarked $25 million for a nonprofit Army venture corporation. The Army named OnPoint, based in Maitland, Florida, on May 7, 2003, to run the Army Venture Capital Initiative, focused on soldier power and energy technologies. Like In-Q-Tel, OnPoint takes equity stakes.
Defense Innovation Unit (DoD, August 2015): contracts, not equity. DIU, originally DIUx, was launched in August 2015 by Defense Secretary Ash Carter in Mountain View, California. It is an adoption and contracting bridge that fast-tracks commercial technology into the Defense Department, typically through prototype and production contracts under Other Transaction Authority. It does not buy equity. That is the clean line between DIU and In-Q-Tel.
Office of Strategic Capital (DoD, December 2022): loans, not equity. DoD established the Office of Strategic Capital in December 2022 to channel private capital into national-security-critical technologies, largely through loans and loan guarantees rather than direct ownership. Paired with the Small Business Administration's Small Business Investment Company Critical Technology initiative, an October 2024 announcement described 12 firms and 13 funds planning to raise about $2.8 billion in private capital matched with SBA-guaranteed loans. That $2.8 billion is targeted private capital to be raised, not $2.8 billion of federal cash handed out.
NASA and DHS: mostly adoption and prize vehicles. NASA and the Department of Homeland Security, through efforts such as the DHS Silicon Valley Innovation Program, run Silicon Valley outreach that is mostly non-equity, using adoption pilots and prize-style awards rather than taking startup stakes.

So of the family, only In-Q-Tel and OnPoint take equity. DIU writes contracts. The Office of Strategic Capital extends credit. NASA and DHS mostly run adoption and prize programs. The shared idea is pulling commercial innovation into government faster than traditional procurement allows. The instruments are not the same, and they do not carry the same transparency or conflict-of-interest exposure.

The honest critique and the honest defence

Both of these are defensible readings of the same record. They are presented side by side because that is what the evidence supports.

The critique. Public money flows through a private nonprofit that is not a government agency and not fully subject to FOIA, so taxpayers get limited visibility into which companies receive funds, on what terms, and what comes back. The 2016 Wall Street Journal investigation found that nearly half of In-Q-Tel's trustees had a financial connection to a funded company, with trustee-linked firms funded on 17 occasions, which raises conflict-of-interest questions even without any finding of wrongdoing. More than 100 investments are classified, so most of the portfolio cannot be audited. The firm's own Form 990 shows CEO compensation near $4 million, disclosed conflict-of-interest transactions, and premium travel. And because gains are reinvested into the mission rather than returned to the Treasury, there is a fair question of whether taxpayers, as the ultimate funders, share in the upside of outcomes like the early Google or Palantir stakes. The model has since been replicated across government without a unified transparency or conflict-of-interest standard binding all of the venture arms together.

The defence. Traditional federal procurement is slow and risk-averse, and it cannot move at startup speed, which effectively locked the intelligence community out of the fastest-moving commercial technology. In-Q-Tel was built to fix precisely that. It is a small, nimble nonprofit that co-invests alongside real venture capitalists, roughly $40 of private money for every $1 it commits, per Fortune, which imposes market discipline and multiplies public dollars rather than replacing private capital. Its stakes are small minority positions, not controlling ownership. It is widely credited with getting the intelligence community access to innovation it could not otherwise reach, from the Keyhole mapping technology that aided troops in Iraq before becoming Google Earth to early bets on Palantir and Anduril. A congressionally mandated BENS panel judged the model "worth the risk" in 2001. And the strongest single argument for the model is that the Army, the Defense Department twice over, NASA, and DHS all built their own versions. Success, In-Q-Tel says, is measured in technologies actually adopted by government users, not in profit.

The record holds both. In-Q-Tel is genuinely credited as a working innovation bridge, and it is genuinely criticised for opacity and insider conflicts. Neither reading cancels the other, and the documented facts, the Form 990s, the Journal investigation, the BENS panel, are what let a reader weigh them.

Fact-check notes and sources

Founding in 1999 under DCI George Tenet, first chairman Norman Augustine, first CEO Gilman Louie, and the "Q" name (via former deputy director Sue Gordon): Wikipedia, In-Q-Tel and In-Q-Tel's own about page.
501(c)(3) status, EIN 52-2149962, and exact total revenue by fiscal year ($184.1M FY2023, $108.26M FY2024, $132.1M FY2025): IRS Form 990 via ProPublica Nonprofit Explorer. The government-only share within each year is reported by In-Q-Tel but treated cautiously here.
CEO Steve Bowsher's reported total compensation (~$3.96M for FY ending March 2025, ~$2.44M base), disclosed conflict-of-interest transactions, and premium travel: IRS Form 990 via ProPublica.
~$100M/yr funding approximation, the ~$1.8B-since-1999 estimate (In-Q-Tel declined to confirm), the $40-per-$1 co-investment ratio, ~1,000 startups screened per year, 800+ companies backed, the Palantir ~$2M stake, and the Bowsher "not the primary goal" quote: Fortune, July 29, 2025. These are Fortune's reporting and estimates, flagged as such.
Keyhole investment (2003), Google's 2004 acquisition, Google Earth's 2005 relaunch, and the caution that the ~$527K figure may span early rounds including Sony: Joe Morrison, industry writer. The ~5,636-share, ~$2.2M Google share sale (November 2005): Wikipedia, In-Q-Tel.
Palantir and CIA-as-sole-customer roughly 2005 to 2008, with Thiel's capital far larger than In-Q-Tel's: Forbes, August 14, 2013.
The 2016 conflict-of-interest findings (nearly half of trustees connected to funded firms; 17 fundings of trustee-linked firms): Wall Street Journal, Damian Paletta, Aug. 30, 2016, via Techmeme aggregation. Reported potential conflicts, not adjudicated wrongdoing. No standalone 2016 GAO report on In-Q-Tel could be located; do not attribute these findings to GAO.
The 2001 "worth the risk" assessment: Business Executives for National Security (BENS) panel report. BENS is a pro-national-security business nonprofit, an advocacy-adjacent source, named here so readers can weigh it.
OnPoint Technologies, Section 8150 of PL 107-117, $25M earmark, and the May 7, 2003 Army announcement (Maitland, FL): U.S. Army Acquisition Support Center.
Defense Innovation Unit launch (August 2015, Ash Carter, contracts not equity): Wikipedia, Defense Innovation Unit.
Office of Strategic Capital (December 2022, loan and loan-guarantee model) and the ~$2.8B private-capital target with SBA: DoD Office of Strategic Capital and the Congressional Research Service overview.
The cross-government landscape of intelligence and military startup investors: CB Insights. The precision that only In-Q-Tel and OnPoint take equity is this piece's own framing of the source material.

In-Q-Tel: How the CIA Built a Venture Fund, and Where the Money Goes