📰 Key Highlights

Apple officially kicked off the 2026 WWDC Developer Conference at Apple Park in California. This year’s conference has two special backgrounds: CEO Tim Cook announced he will officially hand over the reins to Senior VP of Hardware Engineering John Ternus on September 1st, making this his last WWDC; and there’s high external expectation for Apple to catch up in AI, making up for the many gaps accumulated over the past two years in Siri capabilities and core software experience.

The structure of this year’s keynote itself sends a message — Apple chose to lead with “fixes” as the main theme, with feature enhancements coming second. This suggests that issues like user backlash against past design revamps, malfunctioning search features, and frequent AirDrop failures have become internal challenges the company can’t ignore.

On the Siri front, Apple officially confirmed bringing in Google Gemini as the underlying support. The enhanced Siri will have stronger conversational abilities, visual intelligence integration, and launch as a standalone App while maintaining cross-app operation. Senior VP Craig Federighi emphasized that “privacy in AI is non-negotiable,” stating that user data is only used for executing the current request and that external experts can verify at any time.

Additionally, researchers found strings like “foldState” and “angleDegrees” in the iOS 27 developer beta files, hinting at the progress of a foldable iPhone. The official announcement is expected to be saved for the September iPhone annual event.

💬 JudyAI Lab Perspective

Apple WWDC leading with “fixes” as the main theme and new features coming second — this opening itself is an industry signal worth paying attention to, more so than any feature announcement.

Siri officially bringing in Google Gemini as the underlying support, with Craig Federighi simultaneously emphasizing “privacy in AI is non-negotiable,” where data is only used for the current request and external experts can verify. This combination tells us one thing: in the AI capability race, “integrating trusted external models” plus “strict privacy architecture” has become a viable path for big platforms, rather than just the end-to-end self-build route. Apple’s accumulated gaps in Siri capabilities and core experience over the past two years also remind us: constantly stacking features, but issues like malfunctioning search and frequent AirDrop failures — these persistent small frictions can quietly erode user trust.

For builders currently polishing AI products, it’s worth asking yourself: “What are the top 3 small glitches users most frequently hit?” Fixing them often does more to retain trust than launching new features.

📅 Source Info

🔗 Further Reading