Apple held WWDC from June 8 to 12, 2026. The annual developer event carried particular significance this year. Amid the generative AI wave sweeping the industry, it was where Apple finally presented its own answer. It was also the last keynote for Tim Cook, who will step down as CEO on September 1.
His successor will be John Ternus, who has long led the hardware division. Cook’s statement that technology should be personal, powerful, and easy to use captured this year’s direction well. Put simply, the idea is to let AI work quietly deep within the operating system without making users particularly conscious of it.
Siri AI Rebuilt from Scratch and the Gemini Partnership
At the center was Siri AI, rebuilt from scratch. Whereas Siri previously served as a voice interface for predefined commands, the new Siri has become a conversational assistant built on large language models. Behind it lies a partnership with Google.
Reports put the partnership at roughly $1 billion annually and say its reasoning is powered by a custom Gemini model with around 1.2 trillion parameters. Its ability to understand complex phrasing and respond across a wide range of knowledge has improved substantially.
Its appearance has changed too, moving from a small icon in the corner of the screen to a standalone app that retains conversation history. That history is encrypted and synchronized across all devices, and users can finely adjust speaking speed and expressiveness.
Contextual Understanding That Reads the Screen
Particularly useful in practice is its ability to read what is on screen. For example, while viewing an artist’s merchandise announcement on social media, you now only need to say, “Look up the official website, set a reminder for tomorrow’s release time, and include the store link.”
Siri picks up the context from the screen and completes a sequence of tasks across search and calendar operations. In eliminating the need to switch between multiple apps, this feels like the heart of the new experience.
Not every feature is available on every device, however. It was explained that certain capabilities, such as highly expressive speech and advanced dictation, are limited to devices with relatively recent chips.
The Evolution of Generation 27 Operating Systems
The operating systems as a whole have moved to their 27th generation. In response to concerns that the transparency of Liquid Glass, introduced last year, reduced readability, iOS 27 adds a slider for adjusting transparency yourself.
Less visibly, the system-wide search foundation has also been rebuilt, making searches across Spotlight, Mail, and Photos faster and more accurate. Child-protection mechanisms have been significantly strengthened too. Parents can require approval for installing apps and visiting websites and define in detail which apps can be used at different times.
The Photos app adds features that use AI to infer and extend missing backgrounds and naturally remove unwanted objects.
macOS 27 Golden Gate and the Complete End of Intel Support
What was symbolic about macOS 27 Golden Gate was less what it added than what it left behind. Support for Intel Macs, which had continued since 2006, ended completely. The decision was to standardize on Apple chips rather than keep supporting older models lacking the mechanisms needed for advanced AI processing. It was explained that this made app launches roughly 30% faster and AirDrop transfers up to about 80% faster.
Xcode 27 and Vibe Coding
For developers, the major change was AI integration in Xcode 27. So-called vibe coding has become a reality: natural-language instructions can build an app’s basic structure, and continued conversation can add features and translations.
The flexibility to swap Anthropic’s Claude, Google’s Gemini, and OpenAI models like plugins clearly lowers barriers for creators. This year’s Swift Student Challenge again selected 350 students from 37 countries and regions, with 50 receiving distinguished awards. There is no doubt these evolving tools are encouraging young creators.
Availability Split by Region
There is a downside, however. Siri AI and the new Apple Intelligence launched differently across regions. In the EU, regulatory considerations initially prevented availability on iPhone, iPad, and Apple Watch, leaving them available only on Mac and Vision Pro.
Availability in China was also delayed for regulatory compliance. When technological progress meets national rules, user experiences become fragmented in this way.
Ethics Surrounding the Factual Nature of Photographs
Ethical questions also remain. Features that extend backgrounds or remove unwanted objects are convenient, but have prompted debate over whether the result can still be called a photograph.
Apple has prepared safeguards, such as retaining a record of AI processing in edited images and restricting features for political and news images. Yet the question of where to draw the line between fact and fiction will continue.
Assessment: Implementation Rather Than Invention
Looking back at this year’s WWDC, Apple did not demonstrate the invention of new technology. Instead, it seemed to compete through skillful implementation: making existing technology fit naturally into daily life.
As the 27th-generation operating systems and new Siri AI spread fully from autumn 2026 onward, I believe the experiences in our hands will change gradually but surely.
