AI Assistant expanded with time-saving management capabilities

Earlier this year, we introduced AI Assistant, an action-oriented copilot that enhances productivity and decision-making. Since launch, our customers have experienced significant time savings with its first capability, the reference skill. It provides admins with instant answers to questions about Jamf products without the need to comb through documentation or support articles.

Now we're expanding what AI Assistant can do for IT. We have expanded AI Assistant with two new skills – explain and search skill – designed to make navigating Apple device environments even easier.

The search skill: your data, now accessible

Picture this: Your CISO walks in. "How many devices in our European offices are running outdated browsers?" In the old world, you'd create an Advanced Search, add criteria for Chrome, then Firefox, then Safari, then Edge. You'd Google the latest version for each one. Set up your "is less than" operators. Add location filters. Test the logic. Debug why it's showing zero results (oh, it's "Google Chrome.app", not "Chrome"). Twenty minutes later, assuming you didn't mess up the AND/OR operators, you'd have an answer. Maybe you'd make it to lunch, maybe not.

Just type the question as asked. Seconds later, you have the answer, complete with a breakdown by country and browser version.

But here's what changes the game – you can ask follow-up questions like you're having a conversation. "Which of those devices belongs to executives?" Done. Here's your table. "What apps do these outdated devices have in common?" The AI digs in and finds patterns you'd never spot manually.

At a recent Jamf Nation Live community event a customer noted "Just telling it ‘find Macs without FileVault’ felt way simpler than digging through criteria menus.”

The Search skill delivers:

• Conversational data access – no query language, no regex errors, just questions and answers
• Pattern recognition at scale – the AI spots correlations across thousands of devices instantly
• Real-time compliance answers – turn audit nightmares into confident conversations
• Zero Excel required – get presentation-ready insights without leaving Jamf Pro

The explain skill: decode the mystery

Every Jamf admin has that moment. You're troubleshooting an issue, you find a policy called "Q3_Temp_Fix_DONT_TOUCH," and your heart sinks. The creator left two years ago. There's no documentation. It's scoped to 500 devices. Is it safe to disable?

The explain skill is your archaeological assistant. Point it at that mystery policy and it reverse-engineers the intent, tracing through connected scripts, Smart Groups and extension attributes. Most importantly, it can tell you what will break if you change it.

But it goes beyond troubleshooting. Ask it to 'map out what uses our API authentication script' and watch it identify every policy that depends on it. It even spots problems, like three policies still calling an old version of the script with hardcoded credentials. The explain skill helps:

• Inherited environment archaeology – understand years of configurations in minutes
• Pre-deployment sanity checks – catch problems before they reach a single device
• Automated documentation – generate wikis that match reality
• Knowledge preservation – capture institutional knowledge before it walks out the door

A Jamf community member noted, “The idea that it can walk me through our own legacy policies is huge. Our environment’s a bit of a time capsule.”

This isn't AI for AI's sake. It respects your permissions, runs read-only (for now), and requires opt-in activation. It understands Jamf vocabulary, recognizes common patterns, and most importantly, it knows when to pull in extra context. When you ask about a policy, it automatically checks the associated scripts. When you search for outdated apps on devices, it pulls real-time version data from patch feeds before running the comparison.

You'll spend less time clicking through menus and more time solving problems. Less time in spreadsheets and more time being strategic. Less time decoding the past and more time building the future.

Join the beta

Both skills are now available for beta testing in Jamf Pro 11.17 and later. Look for the new AI Assistant button in your top navigation bar that will take you to Jamf Account to opt into the new skill(s). If Jamf SSO isn’t enabled yet, you’ll be prompted to turn it on first.

AI Assistant Legal FAQs

Can the AI Assistant make changes in Jamf Pro?

No, it is strictly read-only for now. Future write-capabilities will be opt-in and security-controlled.

How is security handled?

Through a defense-in-depth strategy. This approach includes:

• Organization-specific authentication
• Data isolation
• Encrypted communication and storage
• Full audit trail for all AI activations and skill usage
• Robust data retention policies

Where is data processed and stored?

All data is encrypted and securely stored in AWS via Amazon Bedrock. Data never leaves the AWS ecosystem and is never used for training. We retain logs of both customer questions (Inputs) and AI Assistant answers (Outputs) for 30 days.

The explain and search skills may store limited configuration and inventory data based on your queries for 30 days. For example, if you query "What users have chrome installed?" AI Assistant will pull from inventory records to identify those users. Records are not directly retained in logs. Rather, information from the logs might be in a response that is in the logs.

Is the AI Assistant enabled by default?

No, it is disabled by default. Admins must explicitly opt-in to use AI assistant, and all AI capabilities can be enabled or disabled granularly (e.g., Retrieval Augmented Generation based on documentation, performing read-only searches in Jamf Pro, etc.)

Where is my data hosted?

Data from AI Assistant is currently stored in a data center located in the U.S. Additional regions, including EU data centers, will be supported soon.

For more information, please refer to the AI Addendum.