What is ChatGPT Work? ChatGPT Work is OpenAI's new agent inside ChatGPT, launched on July 9, 2026, that accepts a business outcome, gathers information across your connected apps, breaks the job into smaller steps, and completes it independently, returning finished spreadsheets, slides, and documents instead of a chat reply. For commercial real estate teams, ChatGPT Work marks a shift from asking an AI questions to handing it entire back-office workflows. That shift matters because the real bottleneck in most CRE shops is not the analysis itself, it is the hours analysts spend assembling rent rolls, operating statements, and lender packages. To see where this fits in the broader landscape, start with our guide to the best AI tools for commercial real estate.
Key Takeaways
- ChatGPT Work, launched July 9, 2026, is an OpenAI agent that completes multi-step tasks end to end, returning finished spreadsheets, slides, and documents rather than chat answers.
- It connects to email, calendars, Google Drive, SharePoint, Slack, and CRM systems, so it can pull deal data and act inside the tools CRE teams already run.
- The best early CRE use cases are repetitive back-office jobs: rent roll cleanup, lease abstraction, investor reporting, draw packages, and pipeline tracking.
- ChatGPT Work runs on GPT-5.6 with Codex built in, and it is metered differently from standard chat, drawing on a workspace credit pool under token-based pricing.
- Governance is central: admins decide what data and actions the agent can touch, and can require human approval before it edits a file or sends an email.
- JLL research found 92% of CRE occupiers are piloting AI while only 5% report hitting most goals, so choosing the right workflow beats a broad rollout.
ChatGPT Work for Commercial Real Estate, Explained
ChatGPT Work for commercial real estate is best understood as an autonomous junior analyst you assign an outcome to, not a chatbot you trade messages with. You describe the result you want, for example a cleaned rent roll or a quarterly investor update, and the agent plans the steps, pulls source files from your connected systems, and hands back an editable deliverable over minutes to hours. Unlike a standard prompt, it stays with a complex project and produces spreadsheets, slide decks, and documents as finished output.
The engine is GPT-5.6, OpenAI's current frontier model, with Codex technology built in so the agent can write and run code to manipulate data. OpenAI has folded its standalone Codex app into a single desktop app, renaming the older experience ChatGPT Classic, and it shipped ChatGPT Work alongside GPT-Live, its new full-duplex voice model, in the same July 9 release. For CRE users, the practical headline is simple: the same assistant that answered questions last year can now execute the workflow behind the question.
Where ChatGPT Work Fits in a CRE Workflow
ChatGPT Work earns its keep on the repetitive, data-assembly tasks that consume analyst hours but rarely require judgment. Because it connects to Outlook and Gmail, Google Drive, SharePoint, Slack, Microsoft Teams, and CRM platforms, it can gather the inputs a task needs and act inside them rather than waiting for a human to copy and paste. The highest-value CRE applications include:
- Rent roll and T12 cleanup: Normalize messy owner-supplied rent rolls and trailing twelve month statements into a consistent template, then compute net operating income, which is gross revenue minus operating expenses and excludes debt service.
- Lease abstraction: Pull key dates, escalations, and renewal options from PDF leases into a structured summary an analyst can verify.
- Investor reporting: Assemble quarterly updates and distribution schedules from your accounting exports, formatted the same way every quarter.
- Accounts payable and draw packages: Match invoices to budget lines and flag exceptions before approval. Our breakdown of AI accounts payable and invoice processing for CRE covers this pattern in depth.
- Pipeline and comp tracking: Keep a running deal pipeline and comparable-sales sheet updated from email and market sources.
Used this way, ChatGPT Work does not replace an underwriter or asset manager. It compresses the hours of prep in front of a deal so the human spends their time on the cap rate assumptions, the DSCR sizing, and the IRR sensitivities that actually move a decision.
ChatGPT Work vs Microsoft Copilot for CRE Teams
The honest answer is that ChatGPT Work and Microsoft Copilot now overlap heavily, and the right pick depends on where your data already lives. Copilot is embedded across Microsoft 365, so it has a native advantage for firms that run on Outlook, Excel, Teams, and SharePoint. ChatGPT Work counters with a more aggressive agentic model, GPT-5.6 with Codex, and cross-platform connectors that reach beyond a single vendor's stack. Firms weighing the two should read our analysis of the Microsoft Copilot shakeup for CRE AI buyers before committing to either as the standard.
For most CRE teams the decision is not exclusive. Many will run Copilot inside Office and pilot ChatGPT Work for the multi-step, cross-system jobs that a single-app assistant handles poorly. The differentiator is not the chat quality, it is which tool can reliably finish a workflow that touches five systems without a person shepherding every step.
Risks and Governance for CRE Firms
Agentic AI raises the stakes because the tool now takes actions, not just drafts text, so governance has to come before scale. OpenAI positions ChatGPT Work with enterprise controls: admins decide which connected tools and data user groups can reach, a Compliance API gives visibility into agent activity, and sensitive steps such as editing a spreadsheet or sending an email can require human approval first. In OpenAI's workspace agents documentation, the company also describes built-in safeguards against prompt injection from misleading external content, and states that business data is not used to train its models by default.
Those controls address the deployment risk, but not the accuracy risk. An agent that assembles a rent roll can still mis-map a column or fill in a value it should have flagged, and in CRE a wrong number flows straight into NOI, cap rate, and valuation. The discipline that separates results from disappointment is verification: treat every deliverable as a draft a human checks, not a finished answer. That gap is well documented. According to JLL's Global Real Estate Technology Survey, 92% of CRE occupiers are piloting AI, yet only 5% report achieving most of their program goals, and more than 60% say they remain unprepared to scale beyond pilots.
How CRE Teams Should Pilot ChatGPT Work
Start narrow, measure, then expand. Pick one repetitive workflow with a clear before-and-after, such as standardizing rent rolls or building the monthly investor report, and run ChatGPT Work against ten real examples with a human checking every output. Track two numbers: time saved and error rate. If the agent saves meaningful hours and its error rate on checked work is low enough to trust with review, widen it to the next workflow. Pair the rollout with a simple internal playbook so the whole team prompts it consistently; our guide to building an AI knowledge base and SOP assistant for CRE teams shows how to capture that institutional knowledge.
If you are ready to move from experimenting with AI to putting an agent into a real CRE workflow, The AI Consulting Network specializes in exactly this. CRE investors looking for hands-on AI implementation support can reach out to Avi Hacker, J.D. at The AI Consulting Network to scope a pilot that fits your systems and your risk tolerance.
Frequently Asked Questions
Q: What is ChatGPT Work and how is it different from regular ChatGPT?
A: ChatGPT Work is an OpenAI agent, launched July 9, 2026, that completes multi-step tasks on its own rather than only answering questions. It gathers information across your connected apps, breaks a job into steps, and returns finished spreadsheets, slides, or documents, staying with a project for minutes to hours.
Q: How much does ChatGPT Work cost for a CRE firm?
A: ChatGPT Work rolls out first to ChatGPT Pro, Enterprise, and Edu subscribers, then expands to Plus and Business plans. OpenAI meters it differently from a standard chat because tasks run longer, drawing on a workspace credit pool under token-based pricing, with admin spend controls. Budget by workflow volume rather than per seat.
Q: Is ChatGPT Work safe for confidential CRE deal data?
A: It can be, with governance. OpenAI states that business data is not used to train its models by default, and enterprise admins control which data and actions the agent can access, including requiring approval before sensitive steps. As with any AI tool, keep a human in the loop and verify every output that feeds an underwriting or valuation decision.
Q: Can ChatGPT Work replace an underwriter or analyst?
A: No. It automates the data assembly around analysis, not the judgment inside it. The strongest results come from letting the agent prepare rent rolls, reports, and lease abstracts while your team owns the assumptions on cap rate, DSCR, and IRR, then verifies the agent's work before it reaches a decision.