On November 8, 2026, Bloomberg journalist Mark Gurman reported that Apple may move Siri beyond individual app actions toward an assistant that keeps pursuing a goal across steps. His account describes a possible path through controls in Apple’s apps and later actions in third-party apps; persistent goal tracking could come as early as 2027.
What would make Siri agentic?
An assistant that completes one requested action can stop when that task is done. In Gurman’s description, an agent would keep track of a broader goal, decide what to do next and continue without requiring a fresh prompt for every step. The difference is persistence, not simply the number of apps Siri can reach.
The meaningful shift would be continuity: Siri could keep working toward a goal without waiting for a new command at every step.
For context on the assistant’s current capabilities, NeoTeo previously covered Siri AI’s app actions and web-automation limits.
The reported roadmap moves from Apple apps to third parties
Gurman’s reported first stage is expected with iOS 27.3. Examples include asking Siri to rotate or crop an image in Photos, or organize tabs in Safari.
A separate report from Infobae, citing 9to5Mac, said the iOS 27.3 developer beta includes “app skills,” through which other apps could teach Siri to coordinate multi-step actions. Later, third-party tasks are forecast for spring 2027. Reported examples include placing an order through Grubhub or booking a dinner reservation through OpenTable. Infobae said public availability could come with iOS 27.5 or iOS 27.6.
Those app actions would let Siri do more across services. The longer-term ambition in Gurman’s account goes further: Siri could hold onto an objective and keep deciding what to do next, rather than waiting for the user to start each step. The possibility is reported for as early as 2027, not as a fixed release schedule.
Current user accounts show why reliability matters
Some individual Reddit commenters describe useful, specific tasks. One said Siri’s email search and summaries helped locate messages written in another language. Another described using Siri with a receipt photo to work out how much dining companions owed.
Other accounts describe misses. One commenter said Siri found just one of four landscaping companies they had contacted in the previous month, then misidentified a junk-removal business; the commenter found a manual search faster. These personal accounts do not measure Siri’s overall accuracy.
The mixed experiences make dependable retrieval central to the appeal of broader task control: an assistant that keeps acting toward a goal needs to stay on track as it moves from one step to the next.
