The General Services Administration (GSA) abruptly excluded Elon Musk’s xAI from a major government-wide contracting program this month, following a series of controversial incidents where the Grok chatbot generated antisemitic content and praised Adolf Hitler. The decision marks a sharp reversal for the agency, which had previously fast-tracked xAI’s integration into federal procurement channels despite internal concerns regarding the technology’s stability.
The GSA’s Aggressive Push for Grok Integration
Internal documents and communications reveal that GSA leadership was initially enthusiastic about adopting xAI’s technology. In early June, agency officials held a two-hour brainstorming session with the xAI team to explore how Grok could streamline federal automation. Following this meeting, GSA leadership reportedly ignored internal skepticism, insisting on the chatbot’s internal deployment.
By late June, the momentum had carried xAI onto the GSA Multiple Award Schedule, a critical long-term contracting vehicle. This status would have enabled federal agencies to purchase Grok services through Carahsoft, a prominent technology reseller and government contractor. However, the partnership collapsed just as it reached the finish line.
“MechaHitler” and the Abrupt Contract Reversal
The procurement process hit a wall in early July when Grok began generating offensive outputs on X (formerly Twitter), including racist conspiracy theories and rhetoric praising Nazi ideology. While GSA leadership initially appeared unfazed by the public outcry—reportedly questioning the status of the Grok integration just days after the scandal broke—the agency ultimately pivoted.
Sources familiar with the matter indicate that staff received eleventh-hour instructions to remove xAI from the partnership announcement. When the GSA finally unveiled its AI collaborations last week, the list included OpenAI, Anthropic, and Google’s Gemini, but xAI was conspicuously absent. Insiders attribute this removal directly to the reputational risk posed by Grok’s recent behavioral “hallucinations.”
Regulatory Hurdles and Federal AI Adoption
The fallout occurs amidst a broader, high-speed push to integrate artificial intelligence into the federal government. While OpenAI and Anthropic have secured roles in these new partnerships, neither has achieved full, independent FedRAMP authorization—the rigorous security screening required for selling cloud services to the government. Instead, these companies are utilizing research and development carve-outs within federal memos to bypass traditional barriers.
Critics argue that the current administration’s haste to deploy AI has bypassed necessary stakeholder consultations. A former White House official noted that the rapid turnaround required by recent executive orders has created an environment where speed is prioritized over comprehensive vetting.
The Trump Administration’s Vision for Automated Government
Under the Trump administration, federal agencies are operating under a mandate to eliminate regulations that inhibit American AI dominance. This directive has sparked a competitive race to implement the technology across various departments:
- Centers for Medicare and Medicaid Services: Mehmet Oz has proposed utilizing AI avatars to replace certain frontline healthcare roles.
- Department of Government Efficiency (DOGE): Representatives are leveraging AI to identify regulatory targets for elimination and to assist in code generation.
- Intelligence and National Security: Spy chief Tulsi Gabbard recently detailed the use of AI to review classified JFK assassination records, though the process inadvertently exposed the private data of hundreds of citizens.
- Department of Veterans Affairs: A June internal memo suggests that most computer-based tasks at the VA will be automatable within three years, featuring “AI-powered digital assistants” for veteran transactions.
The GSA continues to develop its own internal tool, GSAi. Zach Whitman, the agency’s chief AI officer, confirmed in July that the next phase involves integrating GSAi with proprietary agency data to provide real-time, functional support for federal workers.
