OpenAI Chief Executive Officer Sam Altman has said that a future version of ChatGPT — potentially arriving within six months — could continuously monitor a user's computer screen, meetings, and calls to deliver context-sensitive assistance. Altman outlined the vision in an interview with Cory Levy, founder of Z Fellows, describing an AI assistant that understands the full arc of a user's working life rather than responding to isolated prompts.

What Altman Described

Altman envisioned an assistant that 'watches every meeting you're in, records every call, everything that has the perfect context of your whole life,' using that accumulated context to suggest ideas and carry out tasks when authorised by the user. He was explicit that the system would not make autonomous decisions — it would work alongside the user, not replace their judgement.

Users would retain control over what the assistant could access, with the option to connect it to text messages, email, documents, Slack, and other sources depending on personal preference.

The Startup CEO Scenario

Altman illustrated the concept with a scenario involving a startup chief executive managing an information load beyond what is practically trackable. In that setting, the AI could work in the background as the user drafts a sales pitch or a strategy document — flagging potential errors, drawing on customer history and past conversations, and suggesting real-time improvements to a sales email without waiting to be asked.

'I think we're only one model generation away from this actually being incredibly useful,' Altman said, adding that he hoped the capability would fundamentally change the way he works.

What Already Exists

OpenAI has already introduced features that point in this direction. ChatGPT can currently use information from previous conversations to provide more tailored responses, with users able to manage memory and chat history settings — including turning saved memories off entirely.

The Computer History feature for the ChatGPT desktop app on macOS builds a searchable timeline by securely logging desktop interactions such as clicks, app switching, typing, and keyboard shortcuts. The feature is strictly opt-in, automatically ignores private or incognito browsing windows, and can suggest recurring workflows or automations based on repetitive habits.

Privacy and Control Considerations

Altman's framing centred heavily on user control — the assistant accesses only what the user chooses to share, and acts only when authorised. However, the prospect of an AI that continuously monitors screens and records calls is likely to draw scrutiny from privacy advocates and regulators, particularly in jurisdictions with strict data-protection frameworks. This comes amid a broader industry-wide push toward 'agentic' AI systems that operate proactively rather than reactively.

What Comes Next

Altman suggested the capability is 'one model generation away,' implying a potential rollout within the next six months. No formal product announcement has been made, and the timeline remains subject to development progress. The direction signals OpenAI's intent to shift ChatGPT from a prompt-response tool into a persistent, context-aware work companion.