Forget the bold or hyped approaches. In its most anticipated developer event of the year, WWDC, Apple made several announcements to set the stage for on-device and seamless AI, according to IBM experts.
During WWDC, Apple introduced a new generation of LLMs, developed to enhance the Apple Intelligence features, and launched the Foundation Models framework. With this framework, developers will be able to join a suite of tools and tap into Apple’s on-device foundation language model.
“We crafted these generative models to power the wide range of intelligent features integrated across our platforms,” Ruoming Pang, a Distinguished Engineer at Apple AI/ML, wrote on LinkedIn.
The models have improved tool-use and reasoning capabilities, are faster, more efficient and are built to support 15 languages. They are also optimized for Apple silicon and Private Cloud Compute, and include a 3-billion parameter model.
“Giving the developers access to Apple’s AI models, being able to basically play with these models, fine-tune them—I think that that’s really powerful,” says Kaoutar El Maghraoui, a Principal Research Scientist at IBM Research, in an interview with IBM Think.
Another interesting release for developers: Xcode 26. Available on beta versions, Xcode 26 will help developers connect the LLM of their choice directly in their coding experience to generate code, fix bugs or generate documentation.
Get curated insights on the most important—and intriguing—AI news. Subscribe to our weekly Think newsletter. See the IBM Privacy Statement.
Apple also introduced a new software design, made with a new material, Liquid Glass, that reflects and refracts its surroundings. The goal, according to Apple, will be to help bring a greater focus to content. This new design will be available for the first time on all platforms.
“This is our broadest software design update ever,” said Alan Dye, Vice President of Human Interface Design at Apple. This major redesign should also lay the foundation for future experiences.
Apple also previewed the next major macOS release, MacOS Tahoe 26, which leans into iconic elements of a macOS redesign: a more expressive and personal feel. Personalization across devices will be central. Apple announced Live Activities for iPhone, which will keep users updated on their Mac. All operating systems will now use the year in their name, a choice that emphasizes consistency of the user experience across apps and devices.
“I think what Apple wants you to do is start to slide between devices seamlessly and not feel as if you’re on a different device all [the] time,” says Chris Hay, a Distinguished Engineer at IBM. Developers might use this opportunity to be more creative. “Apple is effectively using layers and AI. And I think we’re going to get smarter, better devices, and better applications.”
Many new Apple Intelligence features were also unveiled at WWDC: Live Translation, Image Playground and updates to visual intelligence. With visual intelligence, users will be able to ask ChatGPT questions about what’s on their screen.
“Last year, we took the first steps on a journey to bring users intelligence that’s helpful, relevant, easy to use and right where users need it, all while protecting their privacy,” said Craig Federighi, Senior Vice President of Software Engineering at Apple. “Now, the models that power Apple Intelligence are becoming more capable and efficient, and we’re integrating features in even more places across each of our operating systems.”
Devices will be able to offer a more contextual approach to AI. “Your device has a whole lot of knowledge about you, and over time, this context is going to just be better on the phone,” says Hay.
Running AI on device is also a huge asset for Apple, and its potential to truly appeal to privacy-focused industries. Apple’s end-to-end hardware control and privacy architecture offers many advantages, says El Maghraoui. With full control of the stack, Apple can enforce on-device AI processing down to the hardware level. A consistent interface across devices helps foster the fast adoption of new technologies.
“This also allows Apple to quickly optimize for new chips, like, for example, the neural engines. This vertical integration gives them a unique advantage in performance, privacy, reliability and also the speed to innovate,” she says. “Apple has a competitive advantage, and it sets them apart from the platform-agnostic competitors.”