TLDR: The General Services Administration (GSA) is facing protests from Ask Sage regarding its $1-a-year agreements with AI providers like OpenAI and Anthropic. Ask Sage, led by former Air Force chief software officer Nicolas Chaillan, alleges these pacts circumvent federal acquisition regulations, lack necessary security authorizations, and could lead to vendor lock-in and misleading costs for agencies. A similar agreement with Google for Gemini is priced at 47 cents.
The General Services Administration (GSA) has come under fire for its innovative, yet controversial, $1-a-year agreements with leading artificial intelligence providers, including OpenAI and Anthropic. These agreements, designed to offer federal agencies access to flagship AI products, are now the subject of formal protests filed by Ask Sage at the Government Accountability Office (GAO). A similar agreement with Google for its Gemini AI is priced at 47 cents annually per agency.
Nicolas Chaillan, CEO of Ask Sage and a former Air Force chief software officer, is spearheading the challenge, arguing that these pacts are fundamentally flawed. Chaillan contends that the agreements circumvent federal acquisition regulations, fail to meet crucial security requirements, and mislead agencies regarding the true costs and capabilities of the AI tools. ‘The federal government has never done contracts like this before and for good reason,’ Chaillan told Washington Technology, highlighting concerns about the lack of clarity beyond the initial one-year term.
Ask Sage’s protests, filed on August 21, 2025, specifically target the agreements with OpenAI and Anthropic. The company asserts that these deals violate commercial item pricing requirements, bypass competition mandates, and are inconsistent with GSA’s own OneGov strategy, which aims for centralized software procurement. A significant point of contention is that GSA did not sign these contracts directly with OpenAI and Anthropic, but rather through the reseller Carahsoft, which Ask Sage claims undermines the OneGov initiative.
Furthermore, the protest raises serious security concerns, alleging that the OpenAI and Anthropic offerings lack FedRAMP authorization. ‘Without FedRAMP authorization, agencies cannot conduct proper risk assessments or ensure adequate security controls are in place,’ the protest filing states. This absence of authorization, according to Ask Sage, effectively bars the Department of Defense and other agencies with mission-critical AI requirements from utilizing these services. Additionally, Ask Sage claims the agreements do not meet International Traffic in Arms Regulations (ITAR) requirements.
Chaillan also points out that while the $1 cost provides access to the user interfaces of OpenAI and Anthropic, agencies will still incur regular fees to access the secure AI models themselves, suggesting a hidden cost that could lead to vendor lock-in.
Also Read:
- Anthropic Establishes National Security Advisory Council to Guide AI Integration in Government
- OpenAI’s 2026 Restructuring Imperiled by Microsoft Dispute, Billions in SoftBank Funding at Risk
These protests emerge just weeks after GSA announced the addition of Anthropic’s Claude, Google’s Gemini, and OpenAI’s ChatGPT to its Multiple Award Schedule (MAS) on August 5, 2025, a move lauded by the Trump Administration as accelerating government access to cutting-edge AI. However, Ask Sage’s actions cast a shadow over these initiatives, demanding greater transparency and adherence to established federal procurement standards in the rapidly evolving landscape of government AI adoption.
