How People Are Using ChatGPT at Work in 2026
ChatGPT is becoming more than a place to ask questions. People are using it to draft documents, develop marketing content, prepare learning materials, write and review code, analyze information, and complete other work-related tasks.
New research published by OpenAI on October 26, 2026, offers a closer look at this shift. The company’s Economic Research team released data covering 165 work activities across nearly 100 countries, helping researchers explore how people use ChatGPT for work and how those patterns change over time.
The research includes activities such as preparing instructional materials, developing marketing content, programming, and analyzing business data. Its data spans the first quarter of 2025 through the third quarter of 2026. Read OpenAI’s research announcement.
For employees, freelancers, students preparing for professional careers, business owners, and content creators, the practical question is straightforward: how can ChatGPT help with real work without sacrificing accuracy, privacy, or human judgment?
Quick Answer: People use ChatGPT at work for activities such as writing and editing, research, practical guidance, technical assistance, coding, content development, and data analysis. OpenAI’s latest research provides a more detailed view of these activities across countries and over time. The most useful approach is to give ChatGPT a clearly defined task, relevant context, specific constraints, and an expected output format—then review the result before using it.
What OpenAI’s 2026 Research Reveals About ChatGPT at Work
OpenAI’s October 26 research release expands its OpenAI Signals project, which publishes aggregated information about how people use ChatGPT.
The new release connects work-related messages to 165 categories of activities based on the O*NET framework, a standardized system for describing work activities. This gives researchers a way to compare patterns in ChatGPT usage with other occupational and economic information.
Here are the key details.
| Research detail | What it means |
|---|---|
| Publication date | October 26, 2026 |
| Work activities | 165 categories |
| Geographic coverage | Nearly 100 countries |
| Time period | Q1 2025 through Q3 2026 |
| Main purpose | Examine how people use ChatGPT for work |
| Example activities | Preparing instructional materials, developing marketing content, programming, and analyzing business data |
| Data source | OpenAI Signals individual-use data |
| Important limitation | Does not include enterprise messages or Codex usage |
Source: OpenAI’s October 2026 research announcement.
The study offers researchers a broader view of the tasks people bring to ChatGPT. It does not mean every worker uses AI, nor does it establish that ChatGPT improves productivity for every task.
OpenAI also notes that work-related usage differs across countries. Those differences may reflect variations in the kinds of work people do, access to AI tools, and other factors.
What the numbers do—and do not—tell us
The research is useful because it goes beyond asking whether people use ChatGPT. It examines the types of work they bring to the tool and how those patterns change over time.
However, the dataset focuses on individual ChatGPT plans, such as Free, Go, Plus, and Pro. It excludes enterprise messages and Codex usage, so it should not be treated as a complete picture of AI use across all workplaces.
The data also does not, by itself, prove that AI caused an increase in productivity, reduced working hours, or improved business performance.
For a broader look at individual usage patterns, explore OpenAI Signals’ individual data.
1. Writing and Editing Documents
Writing is one of the practical ways people can use ChatGPT at work. Depending on the task, it can help prepare a first draft, improve clarity, organize ideas, adjust tone, or shorten a long document.
Common examples include:
- Drafting emails and internal announcements.
- Improving reports and proposals.
- Summarizing notes into a readable document.
- Rewriting technical explanations for a general audience.
- Checking a document for inconsistent terminology.
- Creating an outline before writing a full report.
The important distinction is that ChatGPT can help produce and revise text, but the user remains responsible for the final content.
Example: A workplace writing prompt
Copy-and-use prompt:
Help me draft a professional email about [topic].
Recipient: [describe the recipient]
Purpose: [explain what the email needs to accomplish]
Tone: Clear, professional, and respectful.
Key points: [list the information that must be included]
Length: Keep it concise and easy to scan.
Do not invent dates, promises, figures, or details that I have not provided. If essential information is missing, list the questions I should answer before sending the email.
Why this works: The prompt defines the audience, purpose, tone, required information, and limits. These details give ChatGPT a clearer target than a request such as “write a professional email.”
For more examples, visit Promptenza’s AI Prompt Examples library.
2. Research and Information Gathering
ChatGPT can help people understand unfamiliar topics, identify questions to investigate, organize supplied information, and summarize material.
For example, a communications professional might use it to organize background research before preparing a briefing. A business owner might use it to generate questions to investigate before entering a new market.
Useful tasks include:
- Explaining an unfamiliar subject.
- Creating a research plan.
- Organizing notes from supplied documents.
- Comparing options using criteria the user provides.
- Identifying assumptions that need testing.
- Preparing questions for an expert or meeting.
- Summarizing evidence while identifying gaps.
However, ChatGPT’s response should not automatically be treated as verified research. It may omit important information, misinterpret a source, or present an incorrect claim confidently.
Example: A research prompt
Help me research [topic] for [specific purpose].
- Break the topic into the main questions I need to answer.
- Identify the types of sources that would be suitable for each question.
- Separate established information from assumptions and open questions.
- If browsing or source documents are available, cite the sources supporting important factual claims.
- Flag information that may be outdated or requires independent verification.
- End with a concise summary and a list of unanswered questions.
Do not invent sources, quotations, statistics, or research findings.
Practical tip: Ask ChatGPT to show you what still needs verification instead of asking it to make uncertain information sound more convincing.
3. Marketing and Content Development
Marketing content is another practical work area. ChatGPT can help teams brainstorm campaign ideas, outline articles, prepare draft social posts, refine product descriptions, and adapt messages for different audiences.
A marketer might use it to develop several approaches to a campaign before selecting and refining one.
Possible tasks include:
- Brainstorming campaign concepts.
- Drafting social media captions.
- Creating content calendars.
- Organizing customer questions into content themes.
- Rewriting product descriptions for clarity.
- Preparing alternative headlines.
- Turning a supplied brief into a content outline.
These uses can make the drafting process more structured, but they do not eliminate the need for original thinking, brand knowledge, audience research, or fact-checking.
Example: A marketing content prompt
Act as a content strategist helping me plan content for [business or project].
Audience: [describe the intended audience]
Goal: [awareness, education, product discovery, or another goal]
Topic: [insert topic]
Brand voice: [describe the desired style]
Create five distinct content ideas. For each one, provide:
- The central idea.
- The audience problem it addresses.
- The proposed format.
- A suggested headline or opening.
- The information needed to develop it accurately.
Avoid unsupported claims, exaggerated promises, and repetitive ideas. Do not invent customer testimonials or performance results.
For content creators, the biggest advantage is not simply producing more text. It is using a repeatable process to develop, evaluate, and improve ideas.
4. Coding and Technical Problem-Solving
ChatGPT can assist with programming by explaining code, suggesting implementations, helping interpret error messages, and proposing tests.
Depending on the problem and the available tools, developers may use it to:
- Explain an unfamiliar function.
- Draft a small code example.
- Identify possible causes of an error.
- Suggest test cases.
- Explain differences between two approaches.
- Review code for potential weaknesses.
- Improve documentation.
Generated code still needs to be tested. An answer that looks correct may fail in the user’s environment, contain a bug, or introduce a security problem.
Example: A coding prompt
Help me investigate this programming problem.
Language and version: [insert details]
Environment: [describe the relevant framework, tools, or system]
Expected behavior: [explain what should happen]
Actual behavior: [explain what happens instead]
Error message: [paste the relevant error]
Analyze the likely causes before proposing a solution. Explain the recommended fix, identify assumptions, and suggest tests that would help confirm the result. Do not claim the code has been executed or tested unless it actually has.
When working with private code, credentials, customer information, or company systems, follow the organization’s approved AI and data-handling policies.
5. Data Analysis and Business Reporting
People can also use ChatGPT to help understand datasets and prepare reports, provided they have an appropriate way to supply the information and the relevant analysis capabilities are available.
Potential tasks include:
- Explaining spreadsheet formulas.
- Summarizing a supplied dataset.
- Identifying missing or inconsistent entries.
- Suggesting ways to visualize information.
- Drafting an explanation of a chart.
- Comparing results against user-provided criteria.
- Preparing questions for a business review.
The usefulness of the output depends on the quality of the data, the available tools, and whether the analysis has been checked.
Example: A data-analysis prompt
Help me analyze the following dataset for [business question].
What each column means: [provide definitions]
Time period: [provide dates]
Main question: [describe the decision or question]
Please:
- Inspect the data for missing values, inconsistencies, and obvious limitations.
- Summarize the most relevant patterns.
- Distinguish calculated results from possible explanations.
- Show the calculations or method behind important figures.
- Identify conclusions the dataset cannot support.
- Recommend follow-up checks before decisions are made.
Do not invent missing values or imply that a correlation proves causation.
Important: Do not upload confidential business or customer data unless you are authorized to do so and the tool is approved for that information.
6. Preparing Instructional Materials and Learning Resources
OpenAI’s research identifies preparing instructional materials as one of the work activities included in its analysis.
For people who teach, train colleagues, or create educational resources, ChatGPT can help turn a learning objective into an outline, develop practice questions, or explain a concept at different levels.
Examples include:
- Drafting a lesson outline.
- Creating practice exercises.
- Preparing discussion questions.
- Simplifying complex explanations.
- Organizing training notes.
- Developing a knowledge-check quiz.
- Creating examples for a training session.
The material should be reviewed for factual accuracy, appropriate difficulty, accessibility, and alignment with the intended learning goals.
Example: An instructional-design prompt
Help me design a learning resource about [topic].
Audience: [describe learners]
Learning objective: By the end, learners should be able to [specific outcome].
Time available: [duration]
Create a structured outline with:
- A short introduction.
- Three to five learning sections.
- A practical example.
- An exercise that checks understanding.
- Five review questions with an answer key.
Flag claims that need verification and avoid introducing advanced concepts without explaining them.
This approach helps turn a broad request into a defined learning task.
7. Planning, Organizing, and Everyday Work Tasks
Not every useful application requires technical knowledge. ChatGPT can help organize a project, prepare an agenda, structure a meeting summary, or break a large task into smaller steps.
For example, someone planning a project might ask ChatGPT to organize supplied tasks by priority, identify dependencies, and flag information needed before setting deadlines.
Example: A project-planning prompt
Help me organize this project: [describe the project].
Desired outcome: [explain the goal]
Deadline: [insert date, if known]
Available resources: [list the resources]
Known constraints: [list limitations]
Create a plan that includes:
- The main workstreams.
- The tasks required for each workstream.
- Dependencies between tasks.
- Questions that must be answered before scheduling.
- Risks and possible ways to reduce them.
- A short checklist for reviewing progress.
Do not invent commitments, assume unavailable resources, or assign deadlines without a reasonable basis.
Use the result as a planning draft. Confirm dates, responsibilities, dependencies, and resource availability before relying on it.
What These Use Cases Have in Common
The examples above differ, but the strongest prompts generally provide the same kinds of information.
| Prompt component | What it contributes |
|---|---|
| Role or perspective | Establishes the kind of assistance required |
| Task | Defines what ChatGPT should do |
| Context | Explains the situation and relevant background |
| Constraints | States what to include, avoid, or preserve |
| Output format | Makes the expected result clear |
| Quality checks | Helps identify errors, gaps, and unsupported conclusions |
This is why a specific prompt often performs better than a vague one. It reduces ambiguity and gives the AI a clearer standard against which to produce its response.
Explore Promptenza’s Prompt Anatomy guide for a deeper explanation of these components.
How to Get Better Results From ChatGPT at Work
Knowing the possible use cases is only the beginning. The quality of the request, the available context, and the review process all influence the usefulness of the result.
1. Define the actual task
Instead of writing “help me with marketing,” explain the deliverable you need, the audience, and the goal.
A precise task gives ChatGPT a more useful starting point.
2. Provide relevant context
Include the background information necessary to complete the task. This might include an audience profile, a project brief, a style guide, a sample document, or a description of the problem.
Avoid sharing confidential information unnecessarily.
3. Specify the expected format
Tell ChatGPT whether you need a table, an outline, a short email, a report, a checklist, or a step-by-step explanation.
A defined format can reduce the amount of editing required afterward.
4. Set boundaries
Explain what the AI should not assume or change. For example, you can instruct it to preserve supplied figures, avoid inventing citations, or flag missing information instead of guessing.
5. Ask for a review
Ask ChatGPT to identify unclear statements, missing steps, possible inconsistencies, and claims requiring verification.
Treat this as an additional review—not proof that the output is error-free.
6. Verify before using the result
Check important facts, calculations, code, citations, and recommendations. Apply your organization’s policies before sharing AI-generated work externally or using it to make important decisions.
For more guidance, see How to Improve AI Prompts.
Limitations: What ChatGPT Should Not Be Trusted to Do Automatically
ChatGPT can be useful at work, but several limitations remain important.
It can make mistakes. An answer may contain inaccurate information, faulty calculations, or incomplete reasoning.
It may lack essential context. If the prompt leaves out a key requirement, the response may not match the actual task.
It does not automatically know your organization’s rules. You must supply relevant policies and use approved tools.
It may mishandle sensitive information if used carelessly. Follow applicable data-protection requirements and workplace guidance.
It does not guarantee business results. A well-written proposal, report, or marketing plan is not proof that the underlying strategy will succeed.
It cannot replace professional accountability. Decisions involving significant financial, legal, safety, employment, or other consequences need suitable human review.
The most effective approach is to use ChatGPT for assistance while keeping responsibility for verification and final decisions with the appropriate person.
What the Research Means for Workers and Businesses
OpenAI’s October 2026 research provides a more detailed picture of how individual users bring ChatGPT into a variety of work activities. By connecting usage patterns with standardized work categories, it gives researchers a basis for asking more precise questions about how AI is being used.
For workers, the practical opportunity is to identify repetitive or information-heavy tasks where AI assistance may be useful, then test whether it improves the process.
For businesses, the challenge is broader than encouraging employees to use a chatbot. Organizations also need clear rules for data handling, verification, accountability, and the kinds of tasks that require human approval.
For researchers, the data provides a way to investigate how ChatGPT usage varies across countries and work activities. It does not independently establish the effect of AI on employment, wages, or productivity.
These distinctions matter. Usage data can show where people are applying a technology, but further evidence is needed to establish whether those applications produce better outcomes.
Frequently Asked Questions
How are people using ChatGPT at work in 2026?
People use ChatGPT for tasks including writing and editing, research, practical guidance, content development, coding, technical problem-solving, instructional materials, and data analysis. The exact mix varies by user, country, and work activity.
What did OpenAI announce about ChatGPT work usage in October 2026?
On October 26, 2026, OpenAI published research examining 165 work activities across nearly 100 countries. The data covers quarterly trends from Q1 2025 through Q3 2026 and is designed to help researchers study how people use ChatGPT for work.
Does OpenAI’s research include every business using ChatGPT?
No. The individual-use dataset excludes enterprise messages and Codex usage. It should not be interpreted as a complete account of how all organizations use AI.
Can ChatGPT improve productivity at work?
It may help with particular tasks, such as drafting, summarizing, organizing information, or assisting with code. Whether it improves productivity depends on the task, the quality of the output, the time spent reviewing it, and how it fits into the person’s workflow. The cited usage research alone does not prove a productivity gain.
What is the best way to prompt ChatGPT for work?
Describe the task, provide relevant context, specify the intended audience and output format, set clear constraints, and explain how the result should be checked. Review the output before using it.
Can ChatGPT analyze business data?
It can assist with data analysis when the relevant data and appropriate tools are available. Check calculations, data quality, assumptions, and conclusions. Do not upload confidential information unless you are authorized and the tool is approved for it.
Is ChatGPT safe to use for workplace documents?
Safety depends on the information involved, the tool and account settings, and workplace policies. Avoid sharing sensitive information without authorization, verify important claims, and use approved tools for confidential work.
Where can I find prompts for common workplace tasks?
Start with Promptenza’s AI Prompt Examples, AI Prompt Templates, and AI Prompting Techniques.
Test Your Knowledge: ChatGPT at Work in 2026
1. What was the focus of OpenAI’s October 26, 2026 research release?
A. A guarantee that AI increases every worker’s productivity
B. ChatGPT use across 165 work activities in nearly 100 countries
C. A list of every company using ChatGPT
D. A prediction that all jobs will be automated
2. Which is a practical use of ChatGPT for writing?
A. Sending an unreviewed message automatically
B. Inventing missing facts to complete a report
C. Drafting and revising a document using supplied context
D. Guaranteeing that every statement is correct
3. What should you do before relying on AI-generated business analysis?
A. Assume the calculations are correct
B. Verify the data, calculations, assumptions, and conclusions
C. Ignore missing values
D. Treat correlation as proof of causation
4. Does the cited individual-use research include enterprise messages and Codex usage?
A. Yes, it includes all AI activity everywhere
B. It includes only enterprise messages
C. No, it excludes enterprise messages and Codex usage
D. The research does not specify its scope
5. Which prompt is most likely to give ChatGPT a clear task?
A. “Do my work.”
B. “Write something good.”
C. “Prepare a concise project summary for this audience using the supplied notes, flag missing information, and return the result under these headings.”
D. “Make everything perfect.”
Answers
- B — The release examined 165 work activities across nearly 100 countries.
- C — ChatGPT can help draft and revise text, but the result still needs review.
- B — Data quality and verification are essential for responsible analysis.
- C — The individual-use dataset excludes enterprise messages and Codex usage.
- C — It defines the task, audience, source material, constraints, and output format.
Final Takeaway
How people use ChatGPT at work in 2026 is not limited to asking questions. The technology is being applied to a broad range of tasks, from writing and research to coding, instructional materials, and data analysis.
OpenAI’s latest research gives researchers a more detailed way to examine those patterns, but it should not be mistaken for proof that every use case improves productivity.
For everyday users, the practical lesson is to start with a clearly defined task, supply relevant context, specify the expected result, and verify the output. ChatGPT is most useful when it supports a well-understood workflow rather than replacing careful judgment.
Continue learning with Promptenza: Prompt Anatomy · AI Prompting Techniques · AI Prompt Examples · AI Prompt Templates
