OpenAI has released new admin plugins and APIs that allow enterprises to efficiently manage and control ChatGPT Work and the development tool Codex.
Imagine your entire team is conversing with AI and writing code hundreds of times every day. At first, it was great because it accelerated work, but as time went on, you started to wonder, “Who is doing what?” and “Is our data being protected safely?” It’s similar to a situation where dozens of new employees start working simultaneously, and the team leader has to figure out all of their tasks individually.
As corporate AI adoption has become active, such administrative difficulties have become a natural concern. To solve these concerns for enterprises, OpenAI recently released a new ‘Admin plugin’ for ChatGPT Work and Codex. [Source: OpenAI] It is essentially an ‘admin office’ designed to help companies utilize powerful AI tools with greater peace of mind.
Why Is This Important?
Using AI in a company is a completely different issue from simply using a ‘smart chatbot.’ This is because enterprise AI handles various data, ranging from developers’ code to important company project plans.
Until now, as the influence of AI grew, so did the worries of administrators. It was difficult to verify, “Is my team using AI properly?” or “Are security settings configured correctly?” This measure allows IT administrators to secure overall visibility and directly adjust the AI usage environment in line with our company’s security policies. [Source: InfoWorld] Ultimately, it means that while AI can be used freely, its boundaries will be managed firmly.
In Simple Terms
Let’s compare the ‘Admin plugin’ and the new APIs to our company’s systems.
Think of a filter app you use when taking pictures. The user just needs to make the photo look pretty, but the app developer manages the system by looking at statistical data, such as how many photos the user takes and which features they use the most. The newly introduced management tools are the same.
- Admin Plugin: This is a ‘central control panel’ that allows the IT team to see the entire workspace at a glance. You can handle everything from installation to management in one place, such as who is using AI and how the coding assistant tool, Codex, is integrated. [Source: OpenAI]
- Admin API: This is an ‘assistant’ that handles complex, repetitive tasks. If you previously had to click one by one to add members or modify team information, you can now use the API to automatically process tasks such as generating invitation lists, searching member information, and removing permissions. [Source: OpenAI Help Center]
This analogy is easy: if only the tool called AI was provided before, now an ‘operating system’ has been completed that allows you to optimize and monitor that tool according to the enterprise environment.
Current Situation
Recently, OpenAI has been significantly revamping its enterprise services. First, Codex, which was familiar to developers, is no longer an independent, separate app but has been integrated into ChatGPT as a ‘mode.’ [Source: vc.ru]
Furthermore, ‘ChatGPT Work’ has evolved beyond a simple chatbot to possess features as an ‘AI Agent’ that performs complex, multi-step tasks on its own. [Source: ixbt.com] Enterprise customers can now go beyond simply asking questions and have an environment where they can meticulously configure settings such as coding CLI (Command Line Interface) or IDE (Integrated Development Environment) integration through admin tools. [Source: InfoWorld]
Future Outlook
Companies have now entered an era where ‘how safely and efficiently they manage and use AI’ is just as competitive as ‘how well they use AI.’ These management tools are just the beginning. As the scope for more granular usage analysis, tiered permissions by team, and automation of AI tasks expands, the role of the administrator will shift from ‘control’ to ‘efficient collaboration design.’
If you are an IT administrator, now is the time to consider how to set up and optimize your company’s AI usage environment. This is because what is needed for AI to establish itself as a ‘colleague’ in company work is not just a smart brain, but also a reliable shield.
AI Opinion
The adoption of AI in enterprises is no longer a choice but a necessity. However, true operational efficiency is achieved when safe control mechanisms, like these new management tools, accompany AI usage rather than relying on indiscriminate adoption.
References
-
Introducing the Admin plugin for ChatGPT Work and Codex OpenAI - https://openai.com/index/introducing-admin-plugin/ -
OpenAI превратила ChatGPT в рабочего ИИ-агента: новый… ixbt.com - https://www.ixbt.com/news/2026/07/14/openai-chatgpt-chatgpt-work.html - OpenAI объединила Codex с ChatGPT… — Разработка на vc.ru - https://vc.ru/dev/3027916-openai-obedinila-codex-s-chatgpt-i-dobavila-avtonomnogo-agenta-work-chatgpt-ot-openai
-
ChatGPT Enterprise & Edu - Release Notes OpenAI Help Center - https://help.openai.com/en/articles/10128477-chatgpt-enterprise-edu-release-notes -
OpenAI adds plugin system to Codex to help enterprises govern AI coding agents InfoWorld - https://www.infoworld.com/article/4151214/openai-adds-plugin-system-to-codex-to-help-enterprises-govern-ai-coding-agents.html
- To delete chat history for individual users
- To manage and supervise the AI usage environment within an enterprise
- To directly improve the performance of AI models
- It exists as an independent, separate application
- It has been integrated as a mode within ChatGPT
- It has been reduced to a web browser-only service
- Edit all AI responses
- Automate invitation and member management
- Change users' personal passwords