As organizations increasingly integrate AI technologies into their operations, IT administrators face the challenge of managing the configuration and utilization of AI applications across employee devices. Key applications such as Claude Code, Claude Desktop, and OpenAI Codex, which are central to many workflows, require a robust management system to ensure they are used effectively and securely.
Jamf, renowned for managing and securing Apple devices for over 78,000 organizations, has extended its expertise to AI governance. By leveraging Amazon Bedrock, Jamf AI Governance provides a centralized solution for configuring and managing these AI applications on managed Macs, ensuring seamless integration and security.
This article explores how Jamf AI Governance and Amazon Bedrock work together to deploy, configure, and validate AI applications across a Mac fleet.
How Jamf AI Governance Works with Amazon Bedrock
Applications like Claude Code, Claude Desktop, and OpenAI Codex are designed to run locally, utilizing configuration files for essential settings such as inference provider authentication, Model Context Protocol (MCP) server connections, and observability. For enterprise-wide governance, it’s crucial to control where inference occurs and how each application is configured on the device.
Amazon Bedrock facilitates model inference for these applications via your AWS account, with inference operations executed in selected AWS Regions. Jamf AI Governance enables the configuration of necessary parameters to connect each application to Amazon Bedrock, distributing these settings to your fleet through Declarative Device Management (DDM). This collaboration between Amazon Bedrock and Jamf AI Governance offers a scalable method to manage AI applications while maintaining inference security within AWS.
The following architecture illustrates the interaction between Jamf AI Governance, Managed Mac Endpoint, and Amazon Bedrock:
Figure 1: Jamf AI Governance provides configuration to each Mac, and applications use this configuration to connect to Amazon Bedrock for inference.
Policies can be defined in Jamf AI Governance and deployed via Jamf Blueprints. Jamf delivers these configurations to each device’s operating system through DDM, ensuring that managed settings are resistant to local tampering. Users can then seamlessly use the applications without altering local configuration files, while administrators have the ability to monitor policy scope and deployment status in Jamf AI Governance.
Jamf and Amazon Bedrock AI Governance in Practice
In this section, we examine a practical deployment of Claude Code using Amazon Bedrock. The process involves creating a managed policy, deploying it to managed Macs, and validating the policy’s application. This model is applicable to other supported applications, such as Claude Desktop and OpenAI Codex, reinforcing the flexibility and scalability of Jamf AI Governance.
Before starting, ensure you have completed Jamf’s AI governance prerequisites.
Create a policy for Claude Code on Amazon Bedrock
Policies are created within your Jamf account under AI Governance > AI Policies. Using the Policy Builder, you can configure Amazon Bedrock provider settings, including authentication methods, AWS Regions, and template access.
These policies dictate how Claude Code interacts with Amazon Bedrock for organizational use. For instance, enabling Amazon Bedrock prompt caching in Claude Code can significantly reduce costs by up to 90% and latency by up to 85% for supported models. Additionally, you can configure Claude Code’s behavior, such as effort levels, MCP server access, local folder permissions, sandbox settings, and telemetry.
Figure 2: Configuring Claude Code on Amazon Bedrock
Deploy the policy with Jamf Blueprints
Once configured, the policy can be deployed using Jamf Blueprints to the targeted Mac groups. Jamf uses DDM to place the configuration on users’ devices, ensuring it’s applied before opening Claude Code. This means users can start working immediately with Claude Code without manual configuration.
Figure 3: Opening Claude Code with managed configuration applied
Validate and monitor the deployment
After deployment, Jamf AI Governance enables you to review policy scope and deployment status. Additionally, AI Visibility allows you to track AI applications and activities across your fleet and report on governance evidence, providing comprehensive oversight of AI application management.
Figure 4: AI Visibility and Governance Reporting in Jamf AI Governance
Conclusion
The integration of Jamf AI Governance with Amazon Bedrock offers organizations a powerful solution for managing AI applications. By ensuring inference occurs within the secure confines of AWS and offering configuration via DDM, IT teams can deploy settings and controls across a fleet of Macs efficiently. This approach minimizes the need for manual configuration while maintaining robust policy coverage.
For more insights, read the full blog post here.
About the authors
Camille Persson
Cami is a Senior Account Executive at AWS, partnering with ISVs to drive value creation through cloud adoption and AI integration. She is focused on accelerating partner revenue growth and modernizing platforms at scale. Based in Minneapolis, Cami resides with her husband.
Arun Chandapillai
Arun is a passionate Senior Cloud Architect who helps clients accelerate IT modernization through business-focused cloud adoption strategies. He specializes in building and deploying AI and generative AI solutions, from agentic workflows to production-ready applications, on AWS. Arun is a car enthusiast and passionate speaker, passionate about giving back and believes that “you get back what you give”.
Sofiane Hamiti
Sofian is a technology leader with over 12 years of experience building AI solutions and leading high-performing teams to maximize client outcomes. He is passionate about empowering diverse talents to make a global impact and achieve their professional aspirations.
Antonio Rodriguez
Antonio is a senior technical lead for Generative AI at Amazon Web Services. It helps businesses of all sizes solve challenges, embrace innovation, and create new business opportunities with Amazon Bedrock. Outside of work, he enjoys spending time with his family and playing sports with his friends.
Matt Vlasach
Matt is Senior Vice President of Product and Enterprise Solutions Engineering at Jamf, where he helps shape products and solutions for Apple in the enterprise. He brings deep expertise in device management, identity, networking and security, with a focus on security, scalability and ease of use of enterprise technology.
Josh Stein
Josh is VP of Product Strategy and Security at Jamf. A former cybersecurity founder and NSA developer, he brings experience in offensive and defensive security, focusing on helping organizations protect Apple devices against evolving threats.
Just Kaplan
Jen is VP of Product Marketing at Jamf, where she leads product marketing, branding, design and campaigns. She brings experience across SaaS, cybersecurity, retail and digital, with deep expertise in go-to-market strategy and working to help organizations adopt complex technologies through clear and compelling storytelling.
“`
