The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →U.S. federal agencies buy AI in different ways: as software, as an ongoing service, through contracts, or through other agreements. The process is not simply a technical review followed by a purchase. Under the federal lifecycle described in OMB Memorandum M-25-22, agencies define a mission need, research the market, evaluate proposed capabilities, set contract protections, and continue monitoring after deployment. The details below concern federal agencies; state and local procurement rules are outside this scope.
What an agency is buying—and who starts the acquisition
There is no single federal “AI purchase” model. The acquisition may be agency-directed, beginning with a government-identified need, or vendor-driven, when a supplier introduces a capability. The deliverable may be a software product or a continuing service, and the agreement may be a contract or another form of agreement. GAO’s April 2026 review found agencies used varied approaches through fiscal year 2025; it does not identify one route as universally best.
As an Amazon Associate I earn from qualifying purchases.
These distinctions matter because they change what the agency must evaluate and manage. A product and a service can involve different operating arrangements, while an ongoing service can make continuing access, monitoring, and exit provisions especially important. The acquisition team should compare options on their fit for the mission, not assume that a product, service, or agreement type is inherently safer or more effective.
Recommended Free Tools
How the federal acquisition process works
OMB Memorandum M-25-22, issued in April 2025, describes an AI acquisition lifecycle from needs identification and market research through solicitation, testing and award, contract administration, oversight, and closeout. It calls for acquisition decisions to involve both mission and technical expertise.
1. Define the mission need and foreseeable uses
The agency should convene a cross-functional team and identify the problem the AI is expected to address, its operating context, and foreseeable use cases. The team should consider whether high-impact AI may be involved. That early work gives procurement staff, technical specialists, and mission owners a shared basis for comparing proposals and defining what success would mean.
2. Research the market and plan the acquisition
Market research should compare available capabilities and approaches, including new entrants where appropriate. Agencies can seek demonstrations and, where practicable, test proposed solutions under realistic operating conditions. They should examine both likely benefits and limitations, as well as long-term costs and the practical difficulty of changing suppliers later.
3. Write the solicitation and evaluate proposals
The solicitation should give vendors the relevant use context and explain what documentation the agency needs. Evaluation criteria should connect to measurable mission outcomes rather than broad claims about AI. Where practicable, the agency should test proposed solutions to understand how they perform against its needs and where their limitations lie.
4. Award with terms the agency can enforce
The award should translate evaluation priorities into clear responsibilities, rights, and measures. Relevant terms can cover data use, privacy, intellectual-property and licensing rights, performance measures, pricing transparency, continuing testing, and portability. The agency should address how it can obtain knowledge and assets needed to operate, monitor, or transition the system.
5. Operate, monitor, and close out
Before deploying an AI system on an agency-operated information system, the agency needs the required authorization to operate. After deployment, contract administration should support recurring assessment of performance, risks, and effectiveness. At closeout, the agency should assess value and ongoing costs and arrange transfer of data or derived assets as appropriate.
What AI contract terms protect agency data and options
OMB M-25-22 calls for agencies to establish processes for data ownership and intellectual-property rights, including the scope of licenses and the continued availability of components needed to operate and monitor a system. Contract language should be specific enough that the agency can tell what data it controls, what the supplier may retain, and what rights continue after the agreement ends.
Rank #3
- Limits on data collection and retention: Limit vendor collection and retention to what is reasonably necessary to perform the contract.
- Restrictions on secondary training: The memorandum says contracts should prohibit using non-public agency inputs and outputs to further train publicly or commercially available AI unless the agency explicitly consents, consistent with applicable law.
- Privacy and rights: Define applicable privacy protections and agency access to information needed to assess privacy, civil-rights, and civil-liberties risks.
- Intellectual property and licensing: Specify ownership or license scope and whether the agency can continue using components required for operation and oversight.
- Portability and transition: Provide for data and model portability, knowledge transfer, and rights to code or models developed under contract where appropriate.
- Transparent cost and exit terms: Make pricing understandable and set out transition or closeout arrangements, so the agency can assess ongoing costs and avoid unnecessary dependence on one supplier.
These provisions are not interchangeable. For example, a right to export data does not by itself ensure that the agency has the knowledge, licenses, or other components necessary to keep a system operating. The team should consider how the terms work together for the particular product or service.
Do these 3 things before closing this tab:
1Fix the driver behind crashes, sound loss and screen glitches2Repair Windows errors before they cause bigger problems3Scan for outdated or missing drivers - takes under a minuteWhat oversight continues after award
Winning a contract does not end the agency’s responsibility to determine whether an AI system remains suitable. OMB calls for contracts that enable regular performance, risk, and effectiveness assessments, including independent evaluation using agency-defined data where applicable. Agencies should monitor privacy, civil-rights, and civil-liberties risks and, where practicable, define criteria for ending use.
GAO’s accountability framework offers four complementary checks for procurement and ongoing oversight:
- Governance: Is the system’s purpose clear, and are responsibilities for decisions and oversight assigned?
- Data: Is the data appropriate, and is it handled in line with the agency’s requirements?
- Performance: Is there evidence that the system meets the agency’s intended needs?
- Monitoring: Can the agency detect changes in performance or harms over time?
GAO notes that system inputs and operations may not always be visible to an agency. That makes access for review, third-party assessment, and audit relevant considerations when defining oversight arrangements.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What GAO says agencies still find difficult
GAO’s report Artificial Intelligence Acquisitions: Agencies Should Collect and Apply Lessons Learned to Improve Future Procurements (GAO-26-107859), published April 13, 2026, reviewed 13 acquisitions at the Department of Defense, Department of Homeland Security, General Services Administration, and Department of Veterans Affairs. It reported that federal agencies more than doubled their AI use from 2023 to 2024; this is a change in reported use, not a count of acquisitions.
The report identified challenges including access to technical experts who can evaluate proposals and difficulty understanding AI-related costs. GAO also found the selected agencies were not systematically collecting and sharing acquisition lessons. It recommended that DOD, DHS, GSA, and VA require systematic collection and submission of lessons to the GSA-managed repository. The agencies concurred, and GAO listed the recommendations as open in the report.
The finding points to a practical procurement issue: an agency needs enough technical and mission expertise not only to select a proposal, but also to understand costs, evaluate performance, and apply what it learns to future acquisitions.
How current federal guidance differs from earlier snapshots
OMB issued M-25-22 in April 2025. This is the central government-wide lifecycle guidance described here, not a claim that every agency follows identical acquisition procedures. GAO’s December 2023 report had said OMB had not yet issued government-wide guidance on acquiring and using AI; that was the status at the time, before M-25-22.
Other federal figures are also date-specific. GAO-25-107933 identified 94 government-wide AI-related requirements and 10 executive-branch oversight and advisory groups as of July 2025. Those are counts of requirements and groups, not measures of acquisition volume or effectiveness. An earlier GAO report, GAO-24-105980, said 20 of 23 agencies reported about 1,200 current and planned AI use cases based on fiscal year 2022 inventory reporting; that historical figure should not be read as a current total. Likewise, GAO’s statement that industry invested over $250 billion in AI in 2024 refers to reported industry investment, not federal procurement spending.
Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchWindows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallQuick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
