Apple introduces systemwide dictation with AI-powered corrections in iOS 27

At WWDC 2026 on June 8, Apple announced a new systemwide dictation feature for iOS 27 that works across all apps directly from the keyboard. The feature is powered by Apple Intelligence running on Google’s Gemini model and automatically corrects spelling, punctuation, and capitalization as it transcribes speech — without requiring manual editing afterward.

Unlike previous iOS dictation, which treated spoken text as raw input requiring cleanup, the new system applies contextual corrections during transcription. A sentence spoken with uncertain punctuation or a word misspoken against a known vocabulary gets adjusted before it appears on screen.

The competitive context is notable. Google recently released comparable Gemini-powered dictation through Gboard, while third-party dictation apps such as Wispr Flow, Willow, and Monologue have grown by offering intelligent transcription with filler word removal and automatic formatting. Apple previously restricted these apps’ integration in iOS 26.4, requiring additional activation steps that added friction.

For writers and content creators who use iPhones and iPads as part of their workflow, this changes the trade-off calculation between third-party dictation apps and the built-in keyboard. Apple’s system has inherent advantages in cross-app accessibility and requires no separate subscription. Whether it matches the style-awareness of specialized tools remains to be seen.

The feature ships with iOS 27, which extends to iPhone 11 and later.