Apple opens the door to third-party AI model vendors in latest update

There’s been a new development in how Apple manages its products—and it’s an AI disruption.

Apple debuted new AI capabilities at this year’s Worldwide Developers Conference. The company has always had a “consumer-oriented” approach to its products, said Matt Vlasach, Jamf SVP of enterprise product and solutions engineering, who spoke to IT Brew from the conference on June 9.

Now, they’re expanding their software offerings with an open approach to AI, focused on developer tool suite Xcode.

“They’re working with developers quite a bit to encourage them to develop stuff they were doing with Xcode,” Vlasach said. “They’re doing AI-assisted agentic development with Xcode using Anthropic, OpenAI, or Gemini—your subscription of choice.”

IT Brew reached out to Apple for comment for this story, but did not receive a substantive response by press time.

New day. Opening its platform to third-party model providers is a notable step for Apple, which has a reputation for being insular and protective of its intellectual property. But the rise of AI has complicated that picture. Using the Xcode platform, developers will be able to integrate AI coding agents and plugins into their workflow, as well as validate their work “so they can run autonomously for longer, such as writing and running tests, trying ideas in isolation with Playgrounds, checking visual changes with previews, and interacting with the simulator in the new Device Hub,” Apple claims.

For IT pros, especially those at organizations whose tech stack is Mac-driven, the software news is important. It may lead to changes in how teams manage tech and AI within the organization and, potentially, adjust token spend when building apps and services for macOS and iOS.

“We’re going to see more interesting opportunities when it comes to how an IT team can manage either token spend or how LLM routing is done, but that’s just not here today,” Vlasach said.

Take your time. Most developers are unlikely to lean heavily into the new AI capabilities, at least at first, Vlasach said.

Disabling new tools before you trust the models is standard operating procedure for any developer; pros can then educate themselves on the capabilities and move on from there.

“They do their security research, and then they start to enable it,” Vlasach said.