Kandji offers up its take on AI to Apple IT

At the risk of stating the obvious, artificial intelligence (AI) is coming to enterprise IT. Deployment begins with tools to help admins manage their fleets, fix common faults, and optimize existing assets. 

These tools are also coming to the Apple enterprise world in which Cupertino continues to make important investments. Apple device management vendor Kandji has announced its own AI for IT, called Kai. I spoke with Weldon Dodd, senior vice president of global solutions at Kandji, to find out what this new tool offers Apple IT managers who use the device management system.

What is Kai? 

“It’s an AI-driven, natural-language query interface that makes it easy for IT team members to get answers to questions about what’s going on in their Apple device fleet,” Dodd explained.

The best way to understand this AI tool is as an intelligent machine that works with Kandji’s Prism, which provides insights into device fleets using syntax-based query search. What Kai adds is that admins can ask questions in ordinary language to generate responses or reports as needed. “It’s built to understand what you need,” said Dodd.

To do this, the system uses the information Prism has already gathered, then combines this with the contextual intelligence of Open AI to make that information actionable, accessible, and useful.

The company says its tool is especially useful for less technical team members, as it helps them get answers on the fly, build reports, and learn on the job. You can find out more about Kai here.

Why is the time right for AI in Apple IT?

There are plenty of reasons vendors in the burgeoning Apple enterprise space want to optimize their tools. One big reason is expertise. With every business now becoming a tech company, demand for human resources in terms of tech skills is intense. That means the most experienced admins are in huge demand. AI can help fill the gap, enabling the most highly skilled staff to sweat more complex tasks.

Look at the scale at which demand for Apple-related skills is increasing. One 2023 survey claimed that 76% of US companies with over 1,000 staff are using more Apple devices than before. 

Those two reasons alone — higher demand, strained human resources — make the time ripe for AI productivity tools to invade enterprise IT, which, of course, they are. In the Apple enterprise space, we know Jamf has been working with generative AI (genAI) since last year. As Jamf’s big Apple admins event, JNUC, approaches — I imagine Jamf will have its own AI news to share — Kandji has unveiled its own take.

What about privacy?

Privacy and AI is one thing for consumers, where it matters a lot. But, for industries, it’s of vital importance to ensure information pertaining to an enterprise doesn’t slip out. Kanji’s use of OpenAI might be a red flag for some businesses. I asked Dodd about privacy when it comes to user-owned devices, what barriers are in place between corporate and personal use on managed personal devices, and how is everything secured.

According to Dodd, Kai gets the data about a customer’s fleet just in time. “It has access to that data only when necessary to answer a question,” he said. “Kai uses the same tenant-level security as the rest of Kandji’s products. During a user’s session, Kai cannot access any other tenant’s data.”

He also confirmed that data used is sent to OpenAI using that company’s secure AI. The data is encrypted at rest and in transit, and access to it is strictly controlled. “If AI allows us to make our user’s lives easier and help them be more effective in their jobs, we’ll use it,” he said. “This is a perfect use case for it, and we’ll continue to explore others over time. AI isn’t just there to check a box for marketing.”

What about Apple and AI?

I’m on record pointing out that Apple is now building the world’s most AI-friendly ecosystem across Macs, smartphones, and tablets. Dodd doesn’t entirely agree: “As a consumer brand, Apple will use AI to help its users do everyday things more easily and effectively; in that sense, it isn’t really an ‘AI platform,’” he said.

At the same time, the sheer size of Apple’s user base means it is now bringing AI to the masses in a way no one else can. Many millions of people will be making conscious use of genAI just as soon as iOS 18.1 ships next month — and that base is only going to grow, one Mac, iPhone, or iPad upgrade at a time.

Of course, we’ve all heard the claims that AI is gorging itself on global energy supply. How can anyone ignore reports that, “emissions from in-house data centers of Google, Microsoft, Meta and Apple may be 7.62 times higher than the official tally.”

One way to reduce that demand is to create AI models that work at low power on people’s devices, which, of course, is precisely what Apple Intelligence aspires to do. Many of its new LLM models will work quite happily on Mac, iPhone, or iPad; doing so vastly reduces the power required to run each task. Apple is able to achieve this through its ownership of the hardware, software, and key components in its devices.

“Apple’s vertical integration is a key advantage here,” Dodd said. “It can build the powerful processors that AI requires into its devices. We may see LLMs begin to run some of their compute loads locally on devices; Apple is well positioned for that.”

Please follow me on LinkedIn, Mastodon, or join me in the AppleHolic’s bar & grill group on MeWe.