From Coding Tool to Enterprise Work Platform: How OpenAI Codex Is Changing the Daily Work of Knowledge Workers

5M weekly active users | 40% of new users are non-developers | AI agents are redefining “work”

On June 3, 2026, OpenAI held a live event titled “Intelligence at Work,” announcing three major upgrades to its AI coding agent, Codex: role‑specific plugins, Sites, and Annotations. OpenAI also revealed that Codex’s core capabilities will be integrated directly into the ChatGPT app in the coming weeks—giving nearly 1 billion users worldwide a single interface for both conversational thinking and hands‑on execution.

Codex started as a tool for software developers, but its purpose is evolving rapidly—from a “code assistant for programmers” to an “AI agent that helps everyone get work done.” Over the past month, non‑developers—analysts, marketers, operations staff, designers, researchers, investors, bankers—accounted for 40% of new Codex users. Weekly active users have surpassed 5 million, an 8x increase since the start of this year, with roughly 20% of them being non‑programmers.

These numbers send a clear signal: AI is moving from “a specialist’s tool” to “everyone’s work partner.” This article breaks down Codex’s three new capabilities and how they will reshape workflows across different roles.

1. Six Role‑Specific Plugins: Turning Codex into a Domain Expert

Codex’s six new role‑based plugins mark a key step from a general‑purpose tool to a “role‑adapted” assistant. Each plugin bundles the applications, skills, instructions, and workflows needed for a specific job—covering 62 popular apps and 110 skills in total.

Analytics Plugin

Helps analysts and business teams answer questions with data. Users can explore product and business data, explain changes in key metrics, and create reports and dashboards using tools like Snowflake, Databricks Genie, Hex, and Tableau. What used to require writing SQL, building charts, and drafting reports can now be done through natural language prompts.

Creative Production Plugin

Helps marketing and creative teams turn briefs into reviewable assets. Teams can create campaign boards, produce and iterate on display ad variations, and build product scenes using Figma, Canva, Shutterstock, and more. After a creative brief is entered, Codex can automatically generate mood boards, refine product photography, test different versions, and export results directly into design tools.

Sales Plugin

Helps sales teams integrate customer context into deal‑advancing work. Using Salesforce, HubSpot, Slack, and others, sales reps can identify high‑priority leads, prepare for client meetings, complete follow‑ups, create closing plans, and review deals at risk.

Product Design Plugin

Built to turn early ideas into team‑reviewable prototypes. Teams can explore product directions, review user flows, prototype from live URLs, and make static screens interactive.

Public Equities Plugin

Helps investors understand market and company information. Users can leverage premium data sources like Moody’s, FactSet, S&P, and PitchBook to review earnings, compare companies, track signals, and evaluate whether an investment thesis is strengthening or weakening.

Investment Banking Plugin

Helps bankers turn research and due diligence into client‑ready materials. Teams can prepare pitch books, analyze comparable companies and transactions, and translate due diligence into recommendations using trusted data.

Plugins work out of the box, but teams can also adapt them to their own workflows or build and share custom plugins for internal systems. OpenAI also teased more role‑specific plugins on the way—Corporate Finance, Private Equity, Marketing Strategy, Strategic Consulting, and Legal.

2. Sites: Turn Your Work into an Interactive Website

Work doesn’t always fit neatly into documents, spreadsheets, or slide decks. With Sites, Codex can turn ideas, analysis, or plans into a hosted, interactive website or application that others can explore and use.

Sites break the static limits of traditional documents. You can share a site via a URL with anyone in your workspace, giving the team a shared space to explore work, contribute feedback, track progress, and make decisions together.

Practical examples:

  • Create a site for an upcoming client review, generating an interactive page that pulls together product updates, open issues, usage trends, and next steps for that client.
  • Build a scenario planner from a financial model, letting executives compare different assumptions directly instead of jumping between multiple tabs.
  • Turn a launch checklist into a living work hub where the team can find the latest messaging, milestones, owners, and decisions.

Sites can pull together charts for trends, forms for input, trackers for status, or supporting details for decisions. Teams no longer have to force their work into the limits of one tool or file format—they can create experiences that actually fit the work.

OpenAI is building the Sites ecosystem with partners including Wix, Base44, Replit, Lovable, Figma, and Emergent, with support for more complex builds and deployments. Sites are available today in preview for Business and Enterprise teams.

3. Annotations: “Point, then Change” with AI

After the first version of work comes out, refinement often requires judgment, feedback, and iteration. Annotations are built for exactly that—you can add comments directly on Codex’s output, point to the exact part you want to change, and tell Codex what needs to be fixed.

Developers have already been using annotations to refine code, Markdown files, and content created in Codex’s app browser. Now this in‑place editing expands to more types of work.

Examples feel intuitive:

  • Select the navigation bar in a site and ask Codex to update the font.
  • Highlight a claim in an investment thesis and ask where the data came from.
  • Mark a chart in a slide deck and request clearer labels.

Codex focuses updates exactly on the part you selected, so you can refine work precisely without starting over or redoing parts you already like.

4. ChatGPT and Codex Are Coming Together

One of the biggest announcements: OpenAI will integrate Codex’s core capabilities directly into the ChatGPT app in the coming weeks. Users will no longer have to decide between “which one should I use?”—they’ll get both conversation and execution in a single interface.

ChatGPT handles the entry point and dialogue; Codex actually gets the work done. “You might not work 24/7, but your cloud‑based agent can,” said Alexander Embiricos, OpenAI’s Director of Enterprise Product, during the event.

This integration reflects OpenAI’s accelerating push into the enterprise market. According to Chief Revenue Officer Denise Dresser, enterprise already accounts for 40% of OpenAI’s total revenue and will reach 50% by the end of the year.

5. Codex in the Real World: Already at Work Across Industries

Codex isn’t just a concept—it’s already being used in real work settings.

  • Inside OpenAI, non‑technical teams use Codex to build internal apps, prepare executive materials, create dashboards, and turn creative briefs into brand‑compliant assets.
  • At Zapier, teams use Codex to pull knowledge from Slack, Google Docs, and Coda, then turn that context into retrospectives, incident response plans, and feature tickets.
  • At NVIDIA, researchers are using Codex to speed up experimental workflows—from finding research ideas to writing scripts for machine learning infrastructure.

These examples show that Codex is most useful when it works the way your team works—connecting to the tools you already use and always ready to create whatever you need next.

Final Thoughts

From a coding tool to an agentic work platform spanning roles and industries, Codex’s evolution reflects a major shift in AI application: value isn’t just about what AI can generate, but what it can complete.

The three big updates—role plugins, Sites, and annotations—address three critical needs:

  • Adapting AI to domain‑specific knowledge
  • Turning static outputs into dynamic tools
  • Making iteration precise and controllable

Add deep integration with ChatGPT, and Codex is moving from an “AI tool” to a true “AI colleague.”

When AI no longer needs you to teach it “how to do things,” but instead understands what you want to accomplish and helps you get it done, the very definition of work may quietly change. As OpenAI CEO Sam Altman said in the event, the next focus for the AI industry is proactive, autonomous agents that run 24/7. With Codex’s latest updates, that future is becoming reality—one workflow at a time.

Sources: OpenAI official blog, Caixin, ZhiDong, IT House, and other media reports.