ChatGPT Work could change how you research, analyse, and complete projects

OpenAI’s latest product suggests that the next phase of generative AI will focus on execution rather than conversation. For much of ChatGPT’s existence, its appeal rested on immediacy. Ask a question, request a summary, generate a few ideas, then utilise the useful parts elsewhere. OpenAI now wants to close the gap between asking and doing.

ChatGPT Work, announced alongside GPT-5.6, is designed to handle multi-step assignments across a user’s files, applications, and connected sources. Rather than stopping at advice or a draft, it can research information, analyse materials, and produce finished documents, spreadsheets, presentations, reports, and websites. 

The machinery of delegation

The distinction matters. OpenAI now frames Chat as the conversational layer, Codex as the environment for software development, and Work as the place for projects that require gathering context, navigating several steps and delivering a usable result.

A user might ask it to examine a spreadsheet, explain the largest variances, and produce an executive summary, or draw from calendars, emails, and account notes to prepare a briefing for a customer meeting.

Its capabilities include integrations with platforms such as Gmail, Google Drive, Slack, Microsoft Teams, SharePoint, calendars, and customer relationship management systems. It can also create scheduled tasks, monitor information over time, build interactive sites, and operate supported apps and browser interfaces, although OpenAI says consequential actions remain subject to user approval.

The proposition sounds simple: state the outcome, supply the context and let the system determine the path. Yet, the real shift lies in what OpenAI wants users to entrust to it. 

ChatGPT Work enters territory previously occupied by analysts, executive assistants, junior strategists and project coordinators — roles built around assembling fragmented information and turning it into something coherent.

Singapore’s early wager

Singapore will serve as a particularly revealing market. OpenAI says the country ranks among its strongest globally for weekly ChatGPT use per capita, while usage has risen by over 37 per cent year on year. Singapore also ranks among the top five markets for Codex engagement, with over a third of classified Codex requests involving non-coding work.

That appetite does not settle the harder questions. A system that can access emails, files, and internal applications becomes considerably more useful but also more consequential when it misunderstands context, overlooks a constraint, or acts on unreliable information. 

ChatGPT Work, therefore, represents a wager on delegation. OpenAI is betting that users will move from asking AI what to do to letting it do the work — provided they remain willing to check what comes back.