OpenAI CEO Sam Altman has suggested that the next major step for ChatGPT could be an AI assistant capable of understanding a user’s wider digital life, including what appears on their computer screen, meetings, calls, emails and workplace platforms such as Slack. Speaking to Z Fellows founder Cory Levy, Altman said a system with this kind of continuous context could become genuinely useful within the next six months, although OpenAI has not announced a product or launch timeline that matches his prediction.

Altman described a future where an AI assistant would not simply wait for a prompt before responding. Instead, it could have a much broader picture of what a person is doing and use that information to provide suggestions when needed. He said users would decide which information and services the system could access, allowing the assistant to connect details from different parts of their digital work. For example, it could understand the document someone is preparing while also taking into account a customer conversation, previous emails or a discussion during a meeting. The idea is to reduce the need for users to repeatedly explain context to an AI system.

The vision is particularly aimed at people who deal with large amounts of information every day. Altman gave the example of a startup chief executive who may receive hundreds of customer responses and take part in numerous discussions but does not have the time to examine everything personally. An AI assistant with access to those sources could identify patterns, highlight important feedback or point out something that may have been missed. It could also suggest a different approach to a strategy document, identify a possible mistake or offer to complete a task based on information it has already seen. Altman stressed that such a system would work alongside the user rather than simply taking over important decisions.

The idea of an always-on assistant, however, raises questions about privacy and how much access people would actually be willing to give an AI system. A tool capable of watching a computer screen, recording meetings and calls and connecting information from email and workplace applications would potentially have access to some of the most sensitive parts of a person’s professional and personal life. How that information would be stored, processed and protected would become a major issue if such a system were introduced. Altman’s comments did not amount to an announcement of a finished product, and several technical and policy questions would still have to be addressed before the concept could operate at the scale he described.

Altman believes the technology may be only one model generation away from reaching this level of usefulness. His comments point towards a broader change in the way AI assistants could operate, moving from tools that users actively query to systems that understand ongoing activity and provide help based on context. Whether that happens within six months, as Altman suggested, remains to be seen. For now, the comments offer a glimpse into OpenAI’s direction for AI assistants rather than a confirmed feature or launch date.