OpenAI’s 2025 GDPval release does not show that ChatGPT can replace 44 entire jobs. It evaluates how AI models perform on specific work products—such as a financial analysis or sales brochure—across selected occupations. The distinction matters: completing a bounded task is not the same as taking responsibility for a whole role or proving that workers will be displaced.
What OpenAI released—and what “replace” means here
The headline “OpenAI Releases List of Work Tasks ChatGPT Can Already Replace” refers to a September 30, 2025 Futurism story about GDPval, an OpenAI evaluation of model performance on work tasks. Futurism’s framing is stronger than the benchmark’s conclusion. OpenAI describes GDPval as an early, limited way to assess how models might support people at work, not as a demonstration that a model can independently perform an occupation. Futurism’s report and OpenAI’s GDPval announcement describe the release and its scope.
For a personal-finance reader, the practical question is not simply whether AI can produce a spreadsheet, memo, or client-facing document. It is whether that output is accurate, appropriate to a real situation, and usable without substantial human review. GDPval measures a narrower part of that question: performance on defined deliverables.
Which occupations and tasks are covered?
OpenAI says GDPval’s first version covers 44 occupations in nine industries that together contribute more than 5% of U.S. GDP. It contains 1,320 specialized tasks; 220 of those tasks form an open-sourced gold set used for the reported model comparisons. Tasks are based on real work products or comparable constructed deliverables and may involve documents, slides, diagrams, spreadsheets, or multimedia, often with reference files and context. OpenAI reports more than 14 years of average experience among participating professionals and an average of five expert-review rounds per task. These are company-reported features of the evaluation, not evidence that every task or occupation was sampled representatively. OpenAI’s description of GDPval explains its task and expert selection.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
#1 Best Overall
The occupation list spans varied kinds of work, but it is deliberately weighted toward knowledge work. OpenAI used 2024 U.S. Bureau of Labor Statistics wage and employment data and O*NET task classifications to select occupations, focusing on roles where at least 60% of tasks were classified as not requiring physical work or manual labor. It selected five occupations per industry based on wage and compensation contribution. The list is therefore a selected benchmark set, not a census of U.S. jobs.
Occupations in GDPval’s first version
| Industry | Occupations included |
|---|---|
| Real estate and rental/leasing | Concierges; property, real estate, and community association managers; real estate sales agents; real estate brokers; counter and rental clerks. |
| Government | Recreation workers; compliance officers; first-line supervisors of police and detectives; administrative services managers; child, family, and school social workers. |
| Manufacturing | Mechanical engineers; industrial engineers; buyers and purchasing agents; shipping, receiving, and inventory clerks; first-line supervisors of production and operating workers. |
| Professional, scientific, and technical services | Software developers; lawyers; accountants and auditors; computer and information systems managers; project management specialists. |
| Health care and social assistance | Registered nurses; nurse practitioners; medical and health services managers; first-line supervisors of office and administrative support workers; medical secretaries and administrative assistants. |
| Finance and insurance | Customer service representatives; financial and investment analysts; financial managers; personal financial advisors; securities, commodities, and financial services sales agents. |
| Retail trade | Pharmacists; first-line supervisors of retail sales workers; general and operations managers; private detectives and investigators. |
| Wholesale trade | Sales managers; order clerks; first-line supervisors of non-retail sales workers; wholesale and manufacturing sales representatives for technical/scientific products and for other products. |
| Information | Audio and video technicians; producers and directors; news analysts, reporters, and journalists; film and video editors; editors. |
What the examples show
Futurism highlighted examples including a financial analyst creating a competitor landscape for last-mile delivery, a registered nurse assessing skin-lesion images, and a real estate agent designing a sales brochure. Each is a specific deliverable, not proof that ChatGPT can take over the corresponding job. A financial analyst’s role can include deciding which questions matter, checking data quality, explaining uncertainty, and advising decision-makers; a benchmarked output does not establish that the model can handle all those responsibilities. Futurism’s examples illustrate the task-level framing.
What OpenAI reported about model performance
OpenAI says expert graders blindly compared AI-generated deliverables with human-produced work across the 220-task gold set. In its results, Claude Opus 4.1 performed best overall, while GPT-5 was particularly strong on accuracy. OpenAI also reports that performance more than doubled from GPT-4o to GPT-5. These are OpenAI’s comparisons on its chosen tasks and model versions, not an independent measure of performance across workplaces.
OpenAI further reports that frontier models completed GDPval tasks roughly 100 times faster and 100 times cheaper than industry experts. The speed figure refers to model inference time, and the cost figure to API billing rates; both exclude human oversight, iteration, and workplace integration. They should not be read as a claim that an employer can get equivalent real-world work done at one-hundredth the total cost or time. OpenAI’s results and methodology provide the benchmark context.
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Scan for outdated or missing drivers - takes under a minute3Clear out junk files and repair common Windows errorsWhat GDPval does not establish
GDPval is a one-shot evaluation. OpenAI says it does not test whether a model can build context over time, improve a result through multiple drafts, resolve ambiguous requests, or decide what deliverable is appropriate in a real client situation. Nor does the reported speed-and-cost comparison include the human oversight and operational work needed to use outputs in an organization. OpenAI cautions that “most jobs are more than just a collection of tasks that can be written down.”
- It does not show whole-job replacement. The benchmark evaluates deliverables associated with selected tasks, not end-to-end responsibility for an occupation.
- It does not represent all workers. Its 44 occupations were selected from nine industries, with a focus on work that does not require physical or manual labor.
- It does not predict employment outcomes. Task performance alone cannot establish whether jobs will disappear, change, or grow, or how quickly any change might happen.
- It is a snapshot of particular models. Results for GPT-4o, GPT-5, and Claude Opus 4.1 on the 2025 tasks do not automatically describe later models or current ChatGPT features.
How to interpret this for your work or finances
If your role includes producing a clearly specified document, analysis, or other digital deliverable, GDPval is evidence that AI systems may be able to assist with some tasks of that kind. It is not enough to conclude that your job is replaceable. Consider how much of your work depends on context, judgment, accountability, collaboration, and repeated revision—areas the one-shot design does not measure.
For financial decisions, treat AI-generated analysis as a draft or input to verify rather than a substitute for a qualified professional’s judgment. A benchmark result does not validate a particular recommendation for your income, taxes, investments, insurance, or risk tolerance. Check current product capabilities and access conditions before relying on ChatGPT for a task; OpenAI’s release notes are updated as features and availability change: ChatGPT release notes.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What this release means—and what it doesn’t
GDPval provides a structured, task-level comparison of AI-generated work products with expert work on a selected set of knowledge-work tasks. OpenAI’s reported results suggest that models can produce some specified deliverables quickly, but they do not establish that ChatGPT can replace entire occupations, that its benchmark cost advantage carries over to a workplace, or that a particular worker’s job is at risk.
Free tools Windows power users keep installed
One-click scans. No signup required.
Quick Recap
Best Value
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
