# Federal Agencies Adapt AI Hardware Deployment

Federal agencies are adapting AI deployments across data centers and secure, bandwidth-constrained edge environments as hardware changes rapidly and agencies contend with cross-environment security, orchestration, and unpredictable token costs. Lewis Shepherd, Broadcom’s Federal Chief Technology Officer, recommends zero-trust controls, agency-owned infrastructure for validated models, and open enterprise abstraction layers. The signal describes strategic priorities rather than a specific solicitation or contract award.

- Procurement teams can use these priorities when assessing AI hardware and infrastructure proposals, including how vendors address secure operation across environments, model orchestration, and changing hardware requirements.
- Agencies planning AI deployments can evaluate whether zero-trust controls, agency-owned infrastructure for validated models, and open abstraction layers fit their security and operating needs.
- For industry stakeholders, the challenges point to demand for adaptable AI infrastructure and solutions that address edge connectivity, cross-environment security, and cost management; no specific funding, contract vehicle, or procurement timeline is identified.

**Jurisdictions:** federal
**Industries:** Information Technology
**Topics:** Artificial Intelligence, Cybersecurity
**Published:** October 28, 2026

### Vendors
- Broadcom ()

### Key Quotes
> Zero-trust principles are proving essential to secure AI adoption.
> — Lewis Shepherd, Federal Chief Technology Officer at Broadcom

### Sources
- [How federal agencies are adapting to AI hardware acceleration | FedScoop](https://fedscoop.com/video/how-federal-agencies-are-adapting-to-ai-hardware-acceleration) - FedScoop