Jamf launches AI governance for Mac as usage rises

Jamf has launched AI Governance for Mac, now generally available within Jamf for Mac.

The feature is aimed at IT and security teams that want to identify AI tools in use across managed Mac fleets, apply policy controls, and produce reports for compliance and audit purposes. It operates at the operating system level on macOS and covers both approved and unapproved AI tools.

The launch is intended to address a problem for organisations struggling to track AI use on employee devices as staff adopt desktop applications, command-line tools, and background agents. Jamf argues that network and cloud reporting systems do not provide a full picture when AI software runs directly on Apple Silicon machines as local processes.

At launch, the product supports Claude Code, Claude Desktop, and OpenAI Codex. It can govern model access, tenancy, network permissions, file system controls, MCP server restrictions, and other tool-specific settings.

The system also includes what Jamf describes as a vendor control tracking engine to monitor supported AI platforms for new or changed settings. Policies can be enforced offline and before a user first logs in to an AI agent, creating a default policy baseline from the outset.

Mac focus

Jamf positions the product as a native control plane for enterprise AI on Mac, tying governance into the same endpoint management system administrators already use for Apple devices at work. No additional agent is required because the feature uses Jamf's existing telemetry agent and macOS frameworks.

That gives customers three main functions: visibility into AI applications and so-called shadow AI across a device fleet; controls to define which tools are sanctioned and apply those policies at scale; and reporting designed to give senior technology and security executives a snapshot of AI use.

The reporting can feed into security information and event management systems and support reporting against existing compliance frameworks. Different policy settings can also be applied to different teams.

The release comes as businesses try to balance employee demand for AI tools with governance and security concerns. Jamf cited findings from its own survey saying organisations with deeply integrated AI are 40% more likely to report an incident than those still in the exploration phase.

Industry spending on AI governance is also rising. Jamf cited Gartner figures that put AI governance spending at USD $492 million in 2026 and project it to exceed USD $1 billion by 2030.

Beth Tschida, Chief Executive Officer at Jamf, said the product was designed around how AI software operates on Mac endpoints.

"AI adoption across the enterprise is moving faster than existing technology policies can keep up. Organisations need governance that matches the way AI tools actually operate on Mac. This means visibility into what's running, policy controls enforced directly on the endpoint, and reporting that helps security teams demonstrate compliance. Our AI Governance capability delivers that natively from the same platform customers already trust to manage and secure Apple devices," said Tschida.

Partner tools

Jamf said the governance feature can also be used alongside third-party AI and identity products. One example is deploying an agentic identity system such as Okta for AI Agents to managed Mac devices, with Okta's MCP gateway used as the approved route for AI tools to access company data.

In that model, AI agents would authenticate through Okta in the same way as human users, while requests would be verified, authorised, and logged before reaching internal resources. Administrators could also block users from connecting to other MCP servers or AI tools outside those controls.

Organisations can also set a preferred inference provider, including services such as AWS Bedrock, so AI traffic is routed through approved cloud infrastructure. That reflects a wider effort among companies to control not only which tools staff use, but also where data is processed and how access is recorded.

Eventbrite is among the early users cited by Jamf. Its security team said the attraction was the ability to apply policy across Macs without adding another standalone product.

"Like many organizations, we want to enable teams to use AI tools productively while maintaining appropriate governance and oversight. What impressed us about Jamf's AI Governance was how quickly we could apply policy across our Mac fleet without adding another point solution or creating friction for developers. Having this critical capability built into the same device management platform we already use, really simplifies AI governance for our team," said Lalli.