The surge of AI in the workplace introduces new challenges for IT and security teams. Shadow AI, lack of visibility and inconsistent policies create security risks and require teams to take action — many choosing a blanket block of AI platforms. But as organizations push for AI-driven productivity gains, teams need to find ways that enable AI tools without compromising security.

In this webinar, Matt Benyo, Director of AI Initiatives at Jamf, guides Mac admins on their AI governance journey. And in a fireside chat segment, Antonio Rodriguez, Gen AI Technical Leader at AWS Bedrock, gives us a cloud vendor’s perspective of AI implementation.

Watch the webinar for the full story or read on for a quick summary.

Three questions leadership will ask about AI

Engineering leaders and executives are going to have questions about your AI deployment. But often, configuration is absent, manual and unsustainable, or decentralized. And visibility is lacking, or AI shows up in tools that have already been approved. The constantly changing management rules differ between AI platforms, causing additional challenges. Nonetheless, admins have to answer these questions:

• What AI is running on our fleet?
• How is it configured/what does it do?
• Can we prove it?

Benyo walks us through ways to address these questions.

Fireside chat with AWS Bedrock

Benyo asks Rodriguez these questions:

• What are the blockers organizations face when deploying coding tools?
• What teams are needed to make AI governance decisions?
• Why are some organizations working with cloud providers to access AI models instead of going to AI vendors directly?

Check out the webinar for Rodriguez’s answers!

AI maturity isn’t the same as the size of a deployment.

AI is a relatively new challenge, especially at this speed and scale. Whatever stage of AI governance you’re in, you aren’t falling behind. Starting the conversation is important — your path to AI governance can look like this:

1. See: Determine what’s installed, what launched it, what MCP servers are installed and what invocations are touching dangerous surfaces.
2. Decide: Pick a posture, ideally one that balances security and productivity by blocking the most risky AI behaviors.
3. Prove: Keep a record of who deployed what, what devices received it and so on.

In the webinar, Benyo provides a demo of how Jamf will help admins address each of these steps. He emphasizes that this is all done in Jamf Account, so AI governance leaders don’t have to be familiar with or have access to Jamf Pro.

Getting started

Admins have to get their engineering leaders and executives on board, each with their own objectives. Engineering might focus on getting their developers the tools they need; CISO owns the audits and evidence; and CIOs/CTOs need to be ready to prove their posture to the board. Admins have to “light the fuse,” as Benyo says. He suggests:

1. Bringing the question, not the panic.
2. Leading with the inventory, not opinion.
3. Proposing a low risk, balanced governance option.
4. Addressing leadership’s needs.