US agency approves OpenAI, Google, Anthropic for federal AI vendor list

Audio By Carbonatix

Enlarge

By David Shepardson

WASHINGTON (Reuters) -OpenAI’s ChatGPT, Google’s Gemini and Anthropic’s Claude have been added to a list of approved AI vendors, the U.S. government’s central purchasing arm said on Tuesday, as the Trump administration aims to boost AI’s use by federal agencies.

Tuesday’s approvals by the General Services Administration (GSA) are part of a new AI blueprint released on July 23 that aims to loosen environmental rules and vastly expand AI exports to allies, in a bid to maintain the U.S. edge over China in the technology.

The GSA’s step means the approved AI tools will be available for government agencies to use on a platform with contract terms in place.

Federal agencies will explore “a wide range of AI solutions, from simple research assistants powered by large language models to highly tailored, mission-specific applications,” the GSA said.

The GSA added it is focused on AI models “that prioritize truthfulness, accuracy, transparency, and freedom from ideological bias.”

President Donald Trump has called the AI race the fight that will define the 21st century.

His administration’s AI plan, which includes some 90 recommendations, also calls for the export of U.S. AI software and hardware, and a crackdown on state laws deemed too restrictive to let AI flourish.

It is a marked departure from former President Joe Biden’s “high fence” approach that limited global access to coveted AI chips.

The Biden administration last year required federal agencies using AI to adopt “concrete safeguards” on its use and to monitor, assess and test AI’s impacts on the public.

Biden also signed an executive order aimed at promoting competition, protecting consumers and ensuring AI was not used for misinformation, a measure that was rescinded by Trump.

(Reporting by David Shepardson and Harshita Mary Varghese in Bengaluru;Editing by Helen Popper)