Inside OpenAI’s Strategy to Make AI Assistants Ubiquitous

Alexander Embiricos, who leads OpenAI’s enterprise product team, describes the transition as a shift toward intuitive design rather than raw computational power. By reframing complex matrix calculations as a single, continuous conversation, OpenAI hopes to make AI agents feel like a natural extension of the user’s workflow. The assistant, currently available to Pro-tier subscribers for $100 monthly, integrates directly with calendars, emails, and local computer tools to automate tasks without constant oversight.

While the technology is designed to flag conflicts or analyze outages—as seen in early use cases from drone startups—the product faces significant hurdles regarding timing and user trust. The system requires access to sensitive personal data to remain effective, raising questions about data harvesting similar to those currently surrounding Meta’s Muse. Although OpenAI provides controls to opt out of model training, the assistant demands deep integration into cloud drives and personal files to function optimally.

Scaling these agents across the broader digital ecosystem remains the ultimate challenge. While Embiricos acknowledges that businesses and individuals are not yet fully prepared for a world of autonomous software entities—noting he personally avoids meetings with other people’s AI agents—OpenAI is already negotiating with payment processors to expand the agent's reach. Whether users are comfortable granting an AI direct access to their bank accounts, however, remains the next major friction point in the evolution of personal intelligence.