The Intelligence Community Keeps Buying the Same Data Twice. ODNI Wants a Marketplace Instead.

The Office of the Director of National Intelligence is trying to change how the United States intelligence community buys commercial data — not by negotiating better contracts, but by replacing the contract itself as the organizing unit of the purchase.

The vehicle is the Intelligence Community Data Consortium, or ICDC, an effort run by ODNI's Office of the Open Source Intelligence Executive. ODNI put the solicitation on SAM.gov in April 2025 under the identifier ODNI-OT-25-01, and it contemplates awarding one or more prototype Other Transaction agreements under 10 U.S.C. § 4022, as modified by 50 U.S.C. § 3024(n)(6). The government said it was seeking two skill sets: a consortium manager to acquire and manage commercial data on behalf of the whole community, and a software team to build an unclassified marketplace where analysts can find and use it.

The problem statement in the solicitation is unusually blunt for a federal acquisition document. ODNI wrote that the intelligence community has "a commercial data acquisition duplication problem and a costly commercial data replication and storage problem." Its fragmented, decentralized model for buying commercially available information, or CAI, has produced siloed purchases, duplicative buys in which vendors sell the same or similar data to many customers inside the same government, bulk ingest and replication of that data across systems, and limited sharing once it lands. Attempts to fix this with big enterprise licenses, the document says, produced "exorbitantly priced" platform contracts that are "rarely used."

From a transaction to a market

The conventional federal acquisition sequence is linear and terminal: a requirement is written, a procurement is run, a contract is awarded, and delivery follows. The award is the event. Everything after it is administration.

What ODNI described is closer to a market than a transaction. The consortium manager would function as a clearinghouse for community-wide data requirements, a single point of negotiation with vendors, a compliance vehicle for ensuring use complies with intelligence community guidance and vendor terms, and — the part with the sharpest budget implications — "a mechanism to pool agency funds to achieve economies of scale when pursuing access to targeted commercial data sets."

Alongside it sits a technical requirement that matters at least as much. ODNI asked for a "zero copy" solution that keeps data in place on the vendor's infrastructure to the maximum extent possible, with access through API query, platform access, or bulk transfer, delivered through a government-hosted interface the solicitation identified as icdata.gov. The stated goal is to prevent duplicative purchases and cut replication and cloud costs. The practical effect is that access, not possession, becomes the thing being bought.

Those two changes together lower the cost of trying something. If discovery is separated from purchase, and if usage is metered rather than licensed in bulk up front, an agency can test a dataset before committing to it, and a vendor can be measured on delivered value rather than on the strength of a proposal written before anyone used the product. Competition, in that structure, does not end at award. It recurs every time an analyst chooses one source over another.

The unresolved questions

None of this is self-executing, and the model raises real questions that the solicitation does not fully answer.

Accountability is the first. Traditional procurement is heavily instrumented for oversight precisely because the award is a discrete, protestable, auditable event. A persistent marketplace distributes those decisions across thousands of small ones. Who documents why one vendor's data was selected over another's, and how a competitor contests it, is a governance design problem, not a technical one.

Pooled funding is the second. Aggregating demand is what generates the buying leverage, but appropriations are made to agencies, and agencies answer to different committees. Pooling money across them raises questions about who sets priorities when needs conflict.

Security and privacy are the third, and they cut both ways. Keeping data on vendor systems reduces the sprawl of classified copies, but it also means intelligence users are querying infrastructure the government does not own, which puts weight on access controls, logging, and vendor assurance. Civil liberties groups have raised a related concern: that access-without-collection may sit awkwardly against rules written around the act of acquiring data. That debate predates ICDC. A declassified 2022 ODNI senior advisory group report found the community collects a significant volume of CAI and warned it carries growing risks for U.S. person privacy, and ODNI issued a policy framework in May 2024 directing elements to adopt safeguards for sensitive CAI. The ICDC solicitation explicitly requires privacy and civil liberties best practices, including handling of personally identifiable information and role-based access control.

Commercial vendors are already building toward this architecture. 10xNatSec, the sponsor of this article, offers ORRIX, a governed exchange layer for commercial data access across mission networks that centralizes discovery, entitlements, compliance, and usage tracking while data remains with the source — the same access-not-replication pattern ODNI described.

Why it matters beyond data

The logic in ICDC is not specific to datasets. Software, AI models, sensors, and cyber capabilities share the same characteristics: they change faster than a multi-year contract cycle, they are hard to evaluate on paper, and their value is revealed in use. If a persistent, consumption-based marketplace works for commercial data, the case for extending it is straightforward.

That is the strategic stake. Commercial markets now produce national security-relevant capability faster than the acquisition system was built to absorb it. ICDC is one test of whether the government can close that gap by changing how it buys, rather than what.

As of mid-August 2026, ODNI has not publicly announced an ICDC award, and the icdata.gov interface described in the solicitation is not publicly operational. The idea is documented. The execution is not yet visible.