New federal AI platform has huge unacceptable risks, consumer group says

USAi, a closed platform allowing federal employees to experiment with generative AI tools from OpenAI, Anthropic, Google, and Meta, has been launched, the Trump administration’s General Services Administration announced Thursday.

The agency claims the platform will protect privacy by ensuring government data doesn’t feed back into the companies’ models.

However, Public Citizen, a public interest group, says the program opens the door to risky automation and entrenches the dominance of the largest tech firms inside government.

Under the program, OpenAI, Anthropic, Google, and Meta will sell access to their AI models for $1 over the next year.

These terms may appear like a bargain for taxpayers but in reality function as a market capture strategy, J.B. Branch, big tech accountability advocate for Public Citizen, said in a statement.

“Such contracts risk making federal employees dependent on these companies’ systems, granting them an unfair advantage over smaller AI startups, and further cementing their market dominance,” Branch said.

He said these tools are marketed as making employees’ jobs easier, but agentic AI [a system that can accomplish a specific goal with limited supervision] is largely unregulated and untested in making important decisions like loan approval, Medicare enrollment, or Social Security payments.

“The systems may have biased responses tied to historical data, which is troubling given the Trump administration’s ‘Woke AI’ executive order aiming to keep issues of diversity and equity out of AI,” Branch said.

He said the administration is touting a closed federal AI system for safety, but it’s offering no assurances that these AI decisions will reflect equity for all Americans.

“This is a contradictory and dangerous path,” Branch said. “It risks replacing public-sector judgment with corporate-controlled algorithms while sidelining competition and accountability.”

He said the Trump administration needs to pause this AI expansion until enforceable guardrails are put in place to ensure transparency, fairness, worker protections, and open market competition.

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top