ChatGPT keeps creeping toward becoming your AI personal assistant with new scheduled task controls

The retirement of Pulse is the more telling detail here: it suggests OpenAI ran a parallel experiment, got signal on what users actually wanted from persistent agents, and is now collapsing redundant surface area rather than accumulating features. That kind of deliberate subtraction is rarer than it sounds and often precedes a more aggressive capability push.

This is largely disconnected from recent activity in our archive, as we have no prior coverage to anchor against. That said, this story belongs to a broader competitive thread playing out across the assistant market, where Google, Microsoft, and Anthropic are all racing to own the "always-on" layer of user workflows. The scheduling interface is less about convenience and more about retention: a user who has active recurring tasks running inside ChatGPT has a meaningful switching cost that a stateless chat session does not create.

Watch whether OpenAI expands third-party integrations for scheduled tasks within the next two quarters. If outside services can trigger and receive agent outputs natively, the platform argument becomes real; if it stays walled to OpenAI-controlled data sources, this is a retention feature, not a platform play.