Jamf adds Amazon Bedrock support to AI Governance on Mac

Jamf has added support for Amazon Bedrock to Jamf’s AI Governance, letting IT teams centrally configure AI applications on managed Macs while routing model inference through their AWS accounts and chosen AWS Regions. The integration covers applications such as Claude Code, Claude Desktop, and OpenAI Codex and delivers settings to devices via Declarative Device Management.

How do Jamf and Amazon Bedrock govern AI apps on Mac?

Jamf’s AI Governance defines provider and application settings, then delivers them to managed Macs through Declarative Device Management so users can open approved apps without manual setup. Jamf, trusted by more than 78,000 organizations to manage Apple devices at scale, lets administrators specify Amazon Bedrock authentication, AWS Region and model access in an AI policy so inference runs from the AWS Regions they choose and remains in their AWS account boundary.

The integration controls local configuration files that applications use for provider authentication, Model Context Protocol server connections, and observability. Jamf places the managed configuration on devices before the user launches the application, and policies are resistant to local tampering because they are enforced through DDM.

How do you deploy and validate a policy for an application like Claude Code?

You create a managed policy under AI Governance > AI Policies in the Jamf console, configure Amazon Bedrock provider settings (authentication method, AWS Region, model access), then deploy the policy to target Mac groups using Jamf Blueprints, which delivers configuration via DDM. The same pattern applies to Claude Desktop and OpenAI Codex.

Within the policy builder you can enable features such as Amazon Bedrock prompt caching in Claude Code, which the source says can reduce costs by up to 90 percent and latency by up to 85 percent for supported models. After deployment administrators review policy scope and deployment status in Jamf’s AI Governance and use AI Visibility to see AI applications and activity across the fleet and generate governance reports.

Why it matters

Centralizing provider settings and application controls removes manual local configuration and gives IT a single place to define where inference happens, which is important because Amazon Bedrock processes inference through the organization’s AWS account and Regions. For organizations that must keep inference within their security boundary, this pairing means administrators can prevent unapproved inference endpoints while still giving users immediate access to managed AI tools on their Macs.

Jamf’s use of DDM also addresses tamper-resistance at the device level and pairs governance (policy deployment and scope reporting) with visibility (AI Visibility and governance reports), so compliance teams can collect evidence of coverage without relying on end-user setup.

What to watch

Watch how organizations adopt Bedrock-backed policies across device fleets and whether Jamf publishes uptake metrics for policy coverage in AI Visibility. Also monitor vendor support: Jamf’s examples include Claude Code, Claude Desktop and OpenAI Codex, and the same deployment pattern can be extended to other supported applications.