The Department of Health and Human Services wants to keep its workforce on Anthropic's Claude without running a competition first, a move that would hold off rival AI vendors until a broader enterprise contract is in place. In a special notice posted Monday on SAM.gov, HHS disclosed its intent to sole-source an award to Anthropic "for a limited continuity of services award" while it prepares a competitive procurement, FedScoop reported.
The planned award would maintain existing Claude for Enterprise and Claude for Government services for authorized department personnel until HHS awards a multiple-award blanket purchase agreement, according to ExecutiveGov. The notice is a statement of intent, not a completed contract. It set Oct. 21 as the response deadline and did not include a contract value or a projected award date.
Why HHS Says Only Claude Can Fill the Gap
HHS built its justification around continuity. The department argued that Claude is the only tool that can meet its current needs, which FedScoop listed as chat, drafting, summarization, data-analysis workflows, application access and administrative controls.
"The capability is specific to one provider: Claude is Anthropic’s proprietary model family, and the requirement is continuity of the existing Claude deployment already," the notice said, as quoted by FedScoop.
ExecutiveGov reported that the award would maintain access to Claude's existing deployment within HHS infrastructure and that the services would operate within the department's approved security and data-handling requirements. Those requirements include encryption for data in transit and at rest, U.S. data residency, audit logging and provisions for handling protected health information where eligible. ExecutiveGov's list of covered capabilities also includes coding assistance, agentic and delegated-work functions, and secure API access.
The notice does not spell out which existing agreements the sole-source action would continue. FedScoop noted that it is difficult to determine that from the notice alone, though the department's use of Claude is public: HHS rolled the tool out to its workforce in December 2025 and cited Claude use cases in its 2025 AI inventory.
The Oct. 31 OneGov Expiration Behind the Timing
The notice lands days before Anthropic's discounted agreement through the General Services Administration's OneGov initiative is set to end on Oct. 31. That deal gave agencies access to Claude for Enterprise and Claude for Government for $1, and HHS's departmentwide rollout followed the company's offer, according to both outlets.
FedScoop reported that the Oct. 31 end date is itself an extension of the original OneGov agreement, as it is for other providers that signed similar deals. At least one competitor has already signaled how it will price services once the promotional window closes: OpenAI announced a new plan for consumption-based pricing after its own deal concludes, FedScoop said.
For HHS, the expiration creates a practical problem. A department that made Claude available to all of its employees at a nominal price now needs a contract vehicle to keep that service running at whatever price follows, and its planned competitive vehicle is not yet in place.
Is the Sole-Source Justification Sound?
Jessica Tillipman, associate dean for Government Procurement Law Studies at George Washington University, told FedScoop the notice appears to be an admission that HHS is dependent on Claude. She has previously written about the risk of dependency, or vendor lock-in, in AI procurement.
Tillipman said that if switching to a new vendor is too much of a pain or too cost prohibitive, the agency negotiated poorly despite the warnings. She also observed that the notice describes fairly general chatbot services, and said the department's assessment that only Claude could provide them is a questionable use of sole-sourcing.
"To me, it just reeks of … ‘this is going to be a pain, so let’s just keep it going,’ but that’s not a normal justification," Tillipman said.
HHS and Anthropic did not respond to FedScoop's inquiries.
A Split Picture Across Government
The HHS notice arrives against a complicated backdrop for Anthropic in federal work. FedScoop reported that the Food and Drug Administration, an HHS component, had been using Claude in its Elsa chatbot but began transitioning that tool to Google's Gemini earlier this year. That shift followed the Trump administration's governmentwide ban on Anthropic, which was a response to the company's clash with the Pentagon, according to FedScoop, which cited reporting by Fierce Biotech.
The legal battle in that dispute is ongoing. FedScoop noted that while the Pentagon appears to remain at odds with Anthropic, the company seems to be on decent terms with officials elsewhere in government. The HHS notice shows a civilian department seeking to extend, rather than unwind, its reliance on the company's tools at the departmental level, even as one of its own agencies moved a chatbot to a different model.
What the Enterprise Frontier AI BPA Could Change
HHS framed the Anthropic action as a bridge. In the notice, the department said it eventually plans to buy enterprise AI services through a competitive multiple-award blanket purchase agreement, which it called the "HHS Enterprise Frontier SI Models and Advanced Features multiple-award BPA."
ExecutiveGov described the BPA as the next procurement vehicle for enterprise access to frontier AI models and advanced features, with the interim Anthropic award preserving access to Claude in the meantime. The notice leaves the schedule for that BPA unspecified, and FedScoop reported it was not immediately clear whether the competition was active.
A multiple-award structure would, by design, allow HHS to place orders with more than one provider rather than relying on a single deployment. How long the bridge lasts depends on how quickly HHS can stand up that vehicle, a timeline the department has not disclosed.
What It Means for Contractors
Frontier model developers and the integrators that resell or deploy their tools should track the Enterprise Frontier SI Models BPA closely. It is the vehicle through which HHS says it will compete enterprise AI, and its eventual terms will determine whether Anthropic's incumbent position carries over or whether other providers gain a foothold at the department.
The notice also previews a pattern other agencies may face as $1 OneGov offers expire. Departments that adopted a single AI tool broadly under promotional pricing now need a lawful path to continue it, and a sole-source bridge is one option. Contractors that believe they can meet a stated requirement typically use the response window on a notice of intent to say so, although this notice's Oct. 21 deadline has already passed, and the questions Tillipman raised about generic chatbot capabilities suggest such requirements may draw scrutiny.
For integrators building on AI platforms, the episode underscores the value of portability. Tillipman's lock-in critique points to switching costs as a procurement issue in its own right, which could push agencies to write future AI solicitations with stronger provisions for data export, model interchangeability and exit planning.
