Artificial intelligence is shifting rapidly from passive text answering systems toward autonomous agents capable of executing tasks on behalf of users. Nikhil Gupta, an engineer who previously worked on Apple Intelligence and Siri initiatives at Apple, has built a new AI agent named szn. Slated for rollout in October, the service relies on iMessage as its operational interface, enabling users to delegate routine chores through natural messaging conversations.
Executing Multi-Step Chores Directly in iMessage
Standard AI chatbots generally reply to direct prompts with text explanations, but szn is structured as an active agent that takes real-world action. A user simply texts instructions within iMessage, and the system can navigate external websites, conduct product searches, and complete online shopping orders independently. When a purchase or inquiry demands direct coordination with merchants, the agent can communicate with vendors directly to finalize the transaction, removing the friction of juggling multiple retail apps and websites.
Dedicated Phone Numbers and Virtual Assistant Identity
To act as a genuine proxy for the user, szn receives its own dedicated name, phone number, and email address. With these credentials, the AI assistant can send outbound messages to external contacts and answer incoming replies directly. Operating seamlessly within the iMessage environment, it takes on the practical responsibilities of an executive personal secretary, handling logistical follow-ups without requiring constant manual intervention from the device owner.
Integration Across the Apple Service Ecosystem
While an independent application is planned for future release, every subscriber will receive a distinct, dedicated AI model rather than sharing a consolidated instance. Addressing digital privacy concerns, Nikhil Gupta stated that private user information will not be utilized to train the underlying AI models. The platform is also engineered to coordinate with standard Apple services, including reading and editing shared Apple Notes, accessing location coordinates through Find My, and managing files stored across iCloud, all while keeping the user inside the existing iMessage thread.
Platform Dependence and Policy Considerations
By centering the user experience on chat interactions rather than traditional software dashboards, szn presents a distinct architectural direction in consumer AI tools. However, relying on Apple's proprietary messaging network introduces notable platform risks. Apple has previously moved to restrict third-party software that attempts to bridge or operate on top of iMessage. The long-term viability of the service will depend heavily on how Apple chooses to treat an outside AI agent leveraging iMessage as its primary interface.
