Apple’s Siri AI beta in iOS 27 delivers onscreen task handling but still struggles with precise intent parsing
A first-person review of Apple’s revamped Siri AI in the iOS 27 public beta finds onscreen awareness useful and task completion surprisingly reliable, yet still prone to misinterpreting nuanced requests.
1 source · single source
- Apple’s Siri AI beta in iOS 27 can now act on what’s onscreen and answer questions by searching the web, reducing the need to open apps or browsers.
- A reviewer found Siri correctly identified a band’s performance order from a webpage and added calendar events parsed from emails, but sometimes misparsed phrasing like “these tickets” instead of “tickets to this.”
- Onscreen awareness and natural-language tasking are the headline features, but reliability varies by phrasing and app integration.
Apple released the first public beta of iOS 27 on July 13, 2026, marking the debut of a revamped Siri AI that is opt-in and positioned as a foundational step rather than a finished product.
The reviewer, who has been testing the OS since early June, reports that the new Siri can process onscreen content and execute tasks such as adding calendar events parsed from emails, and answering questions by searching the web and extracting details from webpages.
In one example, the reviewer asked Siri what order bands were playing at a concert; Siri read the event page, searched the web, and correctly reported the lineup, reducing the need to open a browser or switch apps.
The reviewer also asked Siri to add WWDC briefings from emails to the calendar, and the assistant parsed the emails and created six events with correct times, though only to the Apple calendar.
Onscreen awareness was highlighted as the most helpful addition, enabling users to ask about visible content and, in some cases, take direct actions like adding an event or routing to an address shown onscreen.
However, the reviewer notes persistent issues with precise intent parsing. For instance, asking Siri to “remind me to buy these tickets when they go on sale” while viewing a concert page created a generic reminder instead of extracting ticket availability, whereas rephrasing to “buy tickets to this when they go on sale” triggered the correct behavior.
The review also flags inconsistent routing behavior, where commands using “route” often failed while “direct” succeeded, underscoring sensitivity to phrasing and integration gaps across apps.
Overall, the experience suggests Apple’s Siri AI is moving toward a more conversational, intent-first model, but remains a work in progress dependent on developer support and user adaptation to specific phrasing patterns.
- Aug 25, 2026 · Google AI — Blog
Google Search adds AI-powered home decor visualization and shopping tools
Trust79 - Aug 25, 2026 · MIT Technology Review — AI
Humanoid robots take center stage at Shanghai ‘carnival’ amid China’s push for embodied AI
Trust79 - Aug 24, 2026 · MIT Technology Review — AI
Why children outperform AI in language learning remains unresolved
Trust79