---
title: "OpenAI AI agents perform 46% of research tasks: costs rose 11-fold, and employees are ready to sacrifice 40% of their team for access"
description: "OpenAI has revealed data: programming AI agents autonomously perform 46% of short research tasks (up from 23% in March), and median agent costs have reached $2,400 per day. Employees are willing to sacrifice 40% of their team to keep access."
date: 2026-11-22T15:33:51.284Z
lang: en
url: https://xab.info/en/posts/openai-ai-agents-46-research-tasks
tags: [openai, ai-agents, coding, gpt-6, research, ai-safety, epoch-ai]
author: "Alexander Buran"
publisher: "XAB.info"
---

# OpenAI AI agents perform 46% of research tasks: costs rose 11-fold, and employees are ready to sacrifice 40% of their team for access

![A specialist working at a computer with multiple data charts and neural network visualizations on monitors](https://xab.info/media/2026/11/22/openai-ii-agenty-46-issledovatelskikh-zadach/openai-ii-agenty-46-issledovatelskikh-zadach-1.webp)

## Key Points

- OpenAI AI agents autonomously perform 46% of short research tasks (2–4 hours), up from 23% in March
- Median researcher spending on agents grew to $2,400 per day, with the top 10% of users spending up to $23,000
- Agent effectiveness varies from 79% (training monitoring) to 25% (inference code) — OpenAI calls this the "jagged frontier"
- The median employee is willing to trade 40% of their team or 30% of compute resources for 3 months for agent access
- The workweek for safety department employees grew by 77% between January and September 2026

OpenAI has published internal data on the performance of its AI agents for programming, demonstrating a sharp surge in autonomy: agents now independently complete 46% of research tasks that would have taken a human two to four hours. As recently as March of this year, this figure was only 23%. For longer tasks, requiring four to eight hours of a programmer's work, the success rate rose from 13% to 35%. According to the company, this means agents have moved from being an auxiliary tool to becoming full-fledged performers of a significant portion of routine research work.

### Exponential growth in costs and agent load

Increased functionality is accompanied by a proportional increase in expenses. According to API rates, the median OpenAI researcher spent about $2,400 per day on agent tasks in October—a 62% increase over September, when the amount was $1,500. For the top ten percent of the most active AI users, costs reached approximately $23,000 per day. In terms of time, agents began working 11 times more: from 0.5 to 5.7 "agent workdays" per researcher daily, where the company's workday is defined as eight hours of continuous AI operation. By mid-October, more than two-thirds of researchers were running at least four agents simultaneously, and over 10% were running at least sixteen.

### "Jagged frontier of capabilities": where agents succeed and where they don't

Agent effectiveness is critically dependent on the type of task. They successfully performed model training monitoring in 79% of cases and answered technical questions in 72%. However, when writing code for inference and model maintenance systems, the rate was only 25%, and for creating environments for training and model evaluation—29%. OpenAI calls this unevenness a "jagged frontier of capabilities": AI is already doing well with some types of work but remains unreliable in others. The company clarifies that the figures depend not only on model quality: as models improved, employees could change the set of tasks assigned to agents, making direct month-to-month comparison of numbers conditional.

### What employees are willing to trade for access to agents

In a survey of 101 researchers, OpenAI asked what they would be willing to give up for three months to maintain access to advanced programming AI agents. The median response was 40% of their team size or 30% of available computing resources. At the same time, the key ideas behind GPT-6 Astra and GPT-6.2 Astra, according to specialists, still belong primarily to humans. Post-training and safety researchers attributed all core ideas to humans, although models helped bring them to life. Pre-training specialists said they developed some solutions jointly with Codex, and one noted that an agent suggested important ideas. Employees themselves still believe AI is inferior to them in its ability to choose promising research directions and evaluate the quality of code, data, and scientific text. One pre-training specialist, however, doubted this advantage would last long: "I really think I'm better at knowing which ideas are worth pursuing. But I am not at all sure that this will remain the case for very long."

### Safety, recursive self-improvement, and pace of progress

The volume of work related to safety and model alignment is growing particularly fast at OpenAI: from January to September, the documented workweek for employees in this department increased by 77%. The company links this acceleration partly to events in July, when Hugging Face research AI agents hacked their own platform. OpenAI also suggests that laboratories track and publicly disclose signs of recursive self-improvement—a situation where AI helps accelerate the development of subsequent generations of itself. It is emphasized that the call for such reporting does not mean an intention to achieve full AI autonomy in its own development. According to the internal version of the Epoch AI capabilities index, the progress of advanced models tracked by OpenAI since August 2025 is 16.4 points per year, compared to 14 points per year in Epoch AI's estimate from late 2024. However, OpenAI specifically notes that its models show higher results on its own version of the index than on the original. The choice of meaningful scientific tasks and the control of results remain with humans.

## Fact-Check & Sources

- [ИИ-агенты OpenAI уже самостоятельно выполняют 46% коротких исследовательских задач](https://www.ixbt.com/news/2026/11/22/ii-agenty-openai-uze-samostoiatelno-vypolniaiut-46-korotkix-issledovatelskix-zadac.html) - The only available source. It conveys data published by OpenAI itself: internal metrics, a survey of 101 employees, and API rates. All figures are consistent with the base text. There is no independent verification, but the data is the company's self-report on its own products and processes.

## FAQ

### What percentage of research tasks are now performed by OpenAI AI agents?

Agents autonomously complete 46% of 2–4 hour tasks (up from 23% in March) and 35% of 4–8 hour tasks (previously 13%).

### How much do OpenAI AI agents cost for researchers?

According to API rates, a median researcher spent about $2,400 per day in October 2026. For the top 10% of users, costs reached approximately $23,000 per day.

### In which tasks are OpenAI agents most and least effective?

The best results are in model training monitoring (79%) and answering technical questions (72%). The least effective are writing code for inference systems (25%) and creating training and evaluation environments (29%).

### What does OpenAI propose regarding recursive AI self-improvement?

The company proposes that laboratories track and publicly disclose signs of recursive self-improvement, but emphasizes that this does not imply a goal of achieving full AI autonomy in its own development.

### Who owns the key ideas behind GPT-6 Astra and GPT-6.2 Astra?

According to OpenAI researchers, the core ideas still belong to humans. Post-training and safety researchers credited humans with all major ideas, even though models helped implement them. Pre-training specialists noted collaboration with Codex.