To provide for certain requirements relating to cloud, data infrastructure, and foundation model procurement.
Sponsors
Legislative Progress
IntroducedMr. Fallon (for himself, Ms. Jacobs, and Mr. Deluzio) introduced …
Summary
What This Bill Does
The "Protecting AI and Cloud Competition in Defense Act of 2025" changes how the Department of Defense buys artificial intelligence systems, cloud computing services, and data infrastructure. It requires the Pentagon to use competitive bidding for these technologies and prevents large tech companies with existing contracts from using government data to improve their commercial products. The bill also mandates annual public reports on competition in the AI marketplace to ensure small businesses can compete fairly.
Who Benefits and How
Small AI and cloud computing companies benefit significantly because the bill requires DoD to actively reduce barriers that prevent them from competing for defense contracts. Multi-cloud technology providers gain a major advantage since the bill prioritizes systems that allow the government to easily switch between different cloud services, creating new revenue opportunities. Non-traditional defense contractors and tech startups benefit from mandatory "modular open systems" requirements that make it easier to enter the defense market without massive upfront investments.
Who Bears the Burden and How
Large cloud providers like Amazon Web Services, Microsoft Azure, and Google Cloud face the biggest burdens. Companies with DoD contracts worth $50 million or more cannot use government-furnished data to train their commercial AI models without explicit Pentagon approval, and they face fines and contract termination if they violate data protection rules. Major AI companies like OpenAI, Anthropic, and Google DeepMind must keep DoD data completely separate from other data and comply with new Defense Federal Acquisition Regulation Supplement (DFARS) rules. These restrictions eliminate the competitive advantage these large providers currently enjoy from vendor lock-in.
Key Provisions
- Defines "covered providers" as any cloud, AI, or data infrastructure company with $50M+ in DoD contracts over the past 5 years, subjecting them to strict data usage restrictions
- Prohibits covered providers from using government-furnished data to train or improve commercial products without express Pentagon authorization
- Requires competitive procurement for every AI, cloud, and data infrastructure purchase, with explicit requirements to reduce barriers for small businesses and non-traditional contractors
- Mandates prioritization of multi-cloud technology that enables switching between providers, unless doing so is infeasible or threatens national security
- Directs the Chief Digital and AI Office to update DFARS with specific penalties (fines and contract termination) for data protection violations
- Requires annual public reports from 2027-2031 assessing competition, innovation, market concentration, and barriers to entry in the defense AI sector
Evidence Chain:
This summary is derived from the structured analysis below. See "Detailed Analysis" for per-title beneficiaries/burden bearers with clause-level evidence links.
Primary Purpose
Regulates DoD procurement of AI, cloud computing, and data infrastructure to promote competition, prevent vendor lock-in, and protect government data rights.
Policy Domains
Legislative Strategy
"Prevent vendor lock-in and monopolistic practices in DoD AI/cloud procurement while maintaining data security and promoting small business participation"
Likely Beneficiaries
- Small AI and cloud computing businesses
- Non-traditional defense contractors
- Multi-cloud technology providers
- DoD (improved competition and data security)
Likely Burden Bearers
- Large cloud/AI providers with existing DoD contracts (covered providers with $50M+)
- Established defense contractors in AI/cloud space (compliance costs, data usage restrictions)
Bill Structure & Actor Mappings
Who is "The Secretary" in each section?
- "the_chairman"
- → Chairman of the Joint Chiefs of Staff
- "the_assistant"
- → Assistant to the Secretary of Defense for Public Affairs
- "the_secretary"
- → Secretary of Defense
- "the_under_secretary"
- → Under Secretary of Defense for Acquisition and Sustainment
- "component_acquisition_executives"
- → Component acquisition executives within DoD
- "the_chief_digital_and_artificial_intelligence_office"
- → Chief Digital and Artificial Intelligence Office (CDAO)
- "the_chief_digital_and_artificial_intelligence_officer"
- → Chief Digital and Artificial Intelligence Officer
Key Definitions
Terms defined in this bill
Has the meaning given in section 5002 of the National Artificial Intelligence Initiative Act of 2020 (15 U.S.C. 9401)
Has the meaning given in NIST Special Publication 800-145 or any successor document
Any company engaged in the provision, sale, or licensing of cloud computing to customers, including individuals and businesses
Has the meaning given in section 101(a) of title 10, United States Code
Any cloud provider, data infrastructure provider, or foundation model provider that has entered into contracts with DoD totaling at least $50,000,000 in any of the 5 previous fiscal years
The underlying computer, network, and software systems that enable the collection, storage, processing, and analysis of data, including recording, transmitting, transforming, categorizing, integrating, and processing data
Any company engaged in the provision, sale, or licensing of data infrastructure to customers, including individuals and businesses
An AI model that generally uses self-supervision, contains at least 1 billion parameters, and is applicable across wide contexts; OR exhibits/could exhibit high levels of performance at tasks that pose serious risk to security, national economic security, national public health, or safety
Any company engaged in the provision, sale, or licensing of foundation models to customers, including individuals and businesses
Architecture and services that allow data, application, and program portability, usability, and interoperability between multiple cloud providers and between public, private, and edge cloud environments with operational consistency, visibility, and resiliency
We use a combination of our own taxonomy and classification in addition to large language models to assess meaning and potential beneficiaries. High confidence means strong textual evidence. Always verify with the original bill text.
Learn more about our methodology