The advocacy groups’ request comes as the federal government is expediting its procurement of AI tools for federal use. Jonathan Raa / NurPhoto / Getty Images
Advocacy groups ask OMB to axe Grok AI procurement
Citing vulnerabilities and biased outputs in the program, multiple advocacy organizations and nonprofits signed a letter to the Office of Management and Budget asking it to bar Elon Musk’s Grok from federal workflows.
Over 30 consumer advocacy non-profit organizations are protesting the potential federal acquisition of large language model Grok, developed by Elon Musk’s company xAI.
In a letter sent to Office of Management and Budget Director Russell Vought on Thursday, the groups — including Public Citizen, the Center for AI and Digital Policy and the Consumer Federation of America — requested that OMB prohibit the use of Grok within federal operations and workflows. The organizations cite concerns over Grok’s output being unreliable and potentially faulty.
“Grok’s track record, which includes producing responses that are often aligned with specific ideological viewpoints rather than objective facts and operational inconsistencies, makes it a prime example of an LLM that is not suitable for government deployment,” the letter reads.
The organizations specifically highlighted concerns over Grok producing ideologically biased output, as well as its alleged vulnerability to cyberattacks.
“At a time when the integrity of federal systems and public trust in government are paramount, we must ensure that the tools procured meet the highest standards of objectivity, truthfulness, and reliability in addition to the minimum risk management practices already required by OMB guidance,” the letter continues.
The Pentagon signed a $200 million deal with xAI — along with three other AI companies — for the use of Grok in mid July. The company announced on X that same day that it was launching Grok for Government, which would be available through the General Services Administration schedule.
The advocacy groups’ request comes as the federal government is expediting its procurement of AI tools for federal use, notably with the new, streamlined USAi program that aims to facilitate the adoption of AI software across government. Leading models from companies like Meta, OpenAI, Google, and Anthropic were included in the launch.
Concern over the bias produced by a given AI model has been a major talking point in the deployment of AI and machine learning in critical environments that handle sensitive data and personal information.
President Donald Trump signed an executive order in July that aimed to advance federal procurement of “ideologically neutral” and “truth-seeking” AI and forbidding biased “woke” AI from government use.
Amid this debate, experts have questioned how provisions within that order will be implemented and worry that the effort to remove “woke” elements may end up chilling free speech.