Apple is finally ready to show its hand on Siri. According to Bloomberg's Mark Gurman, the tech giant plans to unveil a standalone Siri application and a new "Ask Siri" interface during WWDC 2026. The keynote is set for June 8, kicking off a week of developer sessions.

For engineering teams, the architectural shifts matter most. The updated assistant will reportedly ingest personal context from messages, emails, and notes to fulfill requests. This suggests a sophisticated retrieval system capable of grounding generative outputs in user-specific data. Beyond simple queries, Siri aims to execute actions across third-party applications and perform live web searches, moving closer to agentic workflows than traditional voice commands.

The "Ask Siri" function supports natural language input via text or voice, aligning the experience with modern LLM chatbots rather than rigid instruction sets. This flexibility confirms Apple's pivot toward conversational AI models.

Behind the scenes, Apple has confirmed Google Gemini will power the core intelligence. This partnership marks a significant departure from purely internal model development, likely to meet performance benchmarks after repeated delays. While the redesign process has slipped multiple times in recent months, the iOS 27 and macOS 27 updates remain the target deployment vehicles.

Engineers should watch for details on latency, privacy preservation during data indexing, and how much inference happens on-device versus in the cloud. The integration of external foundational models raises questions about data sovereignty and API dependency. After years of stagnation, the June 8 announcement will clarify whether this overhaul delivers substantive technical progress or just a new coat of paint. The industry needs to see if the latency holds up under real-world load. With Gemini involved, the expectation is higher for reasoning capabilities compared to previous iterations.

Source: Engadget