Ternus outlines Siri AI’s architecture and agentic direction
John Ternus described Siri AI as primarily request-based today, while saying its architecture leaves room to expand its capabilities over time. He called agentic capabilities promising, with security, privacy and user understanding as considerations.

Key takeaways · 4
- 01
Ternus characterized Siri AI as primarily request-based at present.
- 02
He said its orchestrator-driven architecture leaves room for capability expansion over time.
- 03
Apple says its beta supports five operating systems and requires an Apple Intelligence-enabled device set to a supported language.
- 04
Apple says certain expanded feature access will be available for a fee in the future.
What Ternus said about Siri
Ternus described Siri AI as primarily request-based at present, rather than characterizing it as already operating mainly through agentic capabilities.[2] He said Siri AI is built on a modern, orchestrator-driven architecture that leaves room to expand its capabilities over time.[2] He called agentic capabilities promising tools for users.[2] These remarks describe a direction and an architectural basis for potential expansion; the evidence does not specify a timetable or list particular future agentic features.[2]
What the beta can do
Apple describes Siri AI as a more capable, conversational assistant with personal-context understanding, broad world knowledge, onscreen awareness and systemwide app actions.[4] Apple says it can find relevant information in a user’s messages, emails and photos, and help draft an email or edit and share a set of photos.[4] It can also answer questions or take actions related to what is on the screen, and use broad world knowledge to provide up-to-date information on virtually any topic.[3] These are Apple’s descriptions of the beta’s capabilities.[4]
Agentic use comes with conditions
Ternus said Apple wants users to understand what they are getting when they use agentic capabilities, and wants those capabilities to protect security and privacy.[2] He also said Apple sees its devices as suitable places to use all forms of AI.[2] The comments set out considerations for how Apple wants such capabilities to work, rather than detailing specific safeguards or a release schedule.[2] Apple also sees intelligence as another business opportunity, according to Ternus.[2]
Availability and limits
Apple says Siri AI is in beta across iOS 27.2, iPadOS 27.2, macOS 27.2, watchOS 27.2 and visionOS 27.2, and requires an Apple Intelligence-enabled device set to a supported language.[3] Apple says it launched the beta for US English users in September and expanded it to five other languages in October.[1] The service is currently unavailable in the EU on iOS, iPadOS and watchOS, and is not available to users under 13.[3] Apple says expanded access to certain features will be available for a fee in the future.[3]
Professionals assessing Siri AI should distinguish its current request-based characterization from the agentic capabilities Ternus described as promising. Apple’s stated emphasis on user understanding, security and privacy offers useful questions to ask when evaluating future capabilities.
Why it matters
Bring this to your team — one session a day, built for your industry.
kju.ai is built for teams: daily sessions, team analytics, and rollout support for your organization.
For teamsHow this developed
5 November 2026
Ternus outlines Siri AI’s architecture and agentic direction
Sources
- Hello, everyone, it’s John: Apple’s Q4 2026 earnings call transcript – Six Colorssixcolors.com
- John Ternus talks plans to improve Siri AI - 9to5Mac9to5mac.com
- Siri AI is available in more languages starting today - Appleapple.com
- Siri AI, a profoundly more capable and personal assistant, is here - Appleapple.com