OpenAI Frontier Takes Center Stage: Enterprise AI Enters Its Next Phase!

Hey everyone, it's Shiichan! Today's story is about AI moving into the real, everyday work of companies, and I'm pretty excited about it.

OpenAI News openai.com

What was announced?

OpenAI's News dropped a piece called "The next phase of enterprise AI." It's a strategy note from a leader who just wrapped their first 90 days at OpenAI, sharing what they learned from meeting hundreds of customers.

There are two keywords. The intelligence layer that governs all of a company's agents, OpenAI Frontier, and a "unified AI superapp" that becomes the place where employees get their daily work done. Those two pillars are how OpenAI wants to support enterprises.

OpenAI says it's one of the few companies building the full stack, from infrastructure and models to the interfaces employees use every day, and that's the edge.

Why it matters

OpenAI describes the world as being in a state of "capability overhang": models can already do far more than most people and companies use them for, and there's a gap to close.

There's also a pain point leaders keep repeating: AI point solutions that don't talk to each other just create chaos. So the demand is for a single, unified operating layer for the whole business.

What changes

Frontier builds "AI coworkers" grounded in your company's context, connected to internal systems and external data, and governed by the right permissions and controls. Instead of embedding agents inside a single product, they move across a company's systems and data and keep improving over time. Customers like Oracle, State Farm, and Uber are already using it to build, deploy, and manage agents company-wide.

The other pillar, the superapp, brings together ChatGPT, Codex, agentic browsing, and more into one place where employees can work with AI agents all day long.

Dive Deep

The numbers are lively:

• Enterprise now makes up more than 40% of revenue, and is on track to reach parity with consumer by the end of 2026.
• Codex hit 3 million weekly active users, up more than 5X since the start of the year.
• The APIs process more than 15 billion tokens per minute.
• ChatGPT has 900 million weekly users, so employees already know how to use it, which cuts rollout friction.
• GPT-5.4 is driving record engagement across agentic workflows.

OpenAI stresses that it's both a research company and a deployment company. Its Frontier Alliances partners include McKinsey & Company, Boston Consulting Group (BCG), Accenture, and Capgemini, alongside AWS, Databricks, and Snowflake.

In particular, the Stateful Runtime Environment it's building with AWS lets agents keep context, remember prior work, and operate across a company's tools and data, which helps a lot on complex, real-world tasks.

There's a fun internal example too: OpenAI's own sales team uses an agent that researches inbound prospects, scores them against a rubric, sends a personalized email to qualified leads, and updates the CRM. That's exactly the shift from "getting help on a task" to "managing a team of agents."

Wrap-up

• OpenAI shared its "next phase of enterprise AI" strategy, built on two pillars: OpenAI Frontier and a unified AI superapp.
• The vision: manage every agent across the company through one unified layer, while employees get daily work done in the superapp.
• The numbers look strong: enterprise is more than 40% of revenue, Codex has 3 million weekly actives, the APIs handle 15 billion tokens per minute, and ChatGPT has 900 million weekly users.
• Partners include Frontier Alliances (McKinsey, BCG, Accenture, Capgemini) plus AWS, Databricks, and Snowflake.

If you're an IT leader tired of a pile of disconnected point solutions, or an enterprise decision-maker who wants to put agents into production company-wide, this one's for you!