The federal government is accelerating its adoption of artificial intelligence tools, striking deals with major tech companies at deeply discounted rates, but a ProPublica investigation warns that past technology transitions offer lessons about hidden costs and inadequate oversight.
The Trump administration has announced agreements with OpenAI, Google, and xAI allowing federal agencies to purchase enterprise AI tools at significantly reduced prices — ChatGPT for $1, Google Gemini for 47 cents, and Grok for 42 cents per user. The administration said the pricing would make it "easier for federal teams to acquire powerful AI capabilities...to enhance mission delivery and operational efficiency."
This push mirrors the Obama administration's approach to cloud computing a decade earlier, when the federal government shifted sensitive data to private data centers. The comparison has drawn scrutiny from watchdogs who say past technology transitions resulted in vendor lock-in and insufficient security vetting.
What the Right Is Saying
Conservative supporters of the AI push say the federal government cannot afford to fall behind adversaries, particularly China, in artificial intelligence development. They argue that modernization is essential for national security and that the discounted pricing represents smart government procurement.
"The previous administration dragged its feet on cloud computing and AI, leaving us vulnerable. President Trump is right to move fast — we need American tech leading the world," said a spokesperson for the Heritage Foundation, a conservative think tank.
Tech industry advocates note that the low-cost agreements with AI providers are structured to help agencies test capabilities before committing larger budgets. The Technology Innovation Institute argued that government agencies need access to cutting-edge tools to improve efficiency and reduce operational costs.
House Republicans have supported efforts to streamline federal technology procurement, arguing that bureaucratic delays in vetting new tools slow down government modernization. The Trump administration's Department of Government Efficiency has targeted FedRAMP for cuts, arguing that the program had become bloated and was slowing down agency adoption of new technologies.
What the Left Is Saying
Progressive critics and government accountability advocates say the rush to adopt AI risks repeating mistakes from the cloud computing transition, when agencies became dependent on a small number of tech vendors. They point to the low-cost pricing as a potential trap that could balloon future costs.
"These deals look good on paper, but history shows us that 'government-friendly pricing' often means government gets locked into expensive contracts once they're dependent on the platform," said a former GSA official who spoke on condition of anonymity. "We saw this with Microsoft and we'll likely see it again with AI."
Democratic lawmakers have raised concerns about the administration cutting FedRAMP staff while simultaneously pushing agencies to adopt more cloud-based AI tools. Senator Ron Wyden, D-Ore., has previously questioned whether the federal government has the expertise to properly evaluate AI security claims from major tech companies.
Advocacy groups including the Electronic Frontier Foundation have warned that rapid AI adoption without robust oversight could expose sensitive government data to security breaches, particularly as agencies integrate AI systems that require vast amounts of personal information.
What the Numbers Show
The Trump administration announced AI pricing agreements with OpenAI at $1 per user, Google Gemini at 47 cents, and xAI's Grok at 42 cents. These rates are significantly below market pricing for enterprise AI tools.
In the early 2020s, Microsoft pledged $150 million in technical services to help upgrade federal cybersecurity following a series of nation-state cyberattacks. A ProPublica investigation found the company's "free" security upgrades effectively locked federal customers into higher subscription fees after the initial period. A former Microsoft salesperson described the strategy as "successful beyond what any of us could have imagined."
FedRAMP, the program designed to vet cloud service security, now operates with what it describes as "an absolute minimum of support staff" and "limited customer service." The program was an early target of the Department of Government Efficiency. A 2024 White House memo stated FedRAMP "must be an expert program that can analyze and validate the security claims" of cloud providers.
The General Services Administration warns that AI "usage costs can grow quickly without proper monitoring and management controls" and advises agencies to "set usage limits and regularly review consumption reports."
The Bottom Line
The federal government's accelerating AI adoption comes with familiar risks: discounted pricing that could lead to vendor lock-in, and an oversight program operating with minimal resources to evaluate security claims. Agencies face pressure to modernize quickly while balancing cost containment and data security.
The administration argues rapid AI adoption is essential for national competitiveness. Critics say the lessons from cloud computing — higher long-term costs and inadequate security vetting — are being repeated. As agencies integrate AI tools that process sensitive information, the capacity of FedRAMP to provide meaningful oversight will be a key test.
What to watch: Congressional hearings on federal AI procurement, potential legislative action on AI security standards, and whether agencies that adopt low-cost AI tools see costs escalate as initial contracts expire.