What is Frontier: The next big enterprise bet from OpenAI?

OpenAI has unveiled a new enterprise-focused platform called Frontier, aimed at helping companies deploy AI “workers” that can collaborate with each other and operate in context-rich business environments.

“The companies that succeed in the future are going to make very heavy use of AI. People will manage teams of agents to do very complex things,” CEO Sam Altman said.

Why is Frontier significant?


Frontier is among OpenAI’s first offerings designed specifically for large enterprises. It enables businesses to build, deploy, and manage AI agents that can perform real work across internal systems.

With this, OpenAI is positioning itself more directly against enterprise-focused rivals such as Claude Enterprise, which already offers similar capabilities.

The company says Frontier helps break down data silos, allowing firms to use AI in a more integrated and holistic way rather than in isolated workflows.

How does it work?

Frontier provides AI agents with shared business context, structured onboarding, feedback-driven learning, and defined access controls—mirroring how human employees operate within organisations.

The platform integrates with data warehouses, CRM systems, ticketing tools, and internal applications to improve contextual understanding. Agents can handle complex tasks such as working with files, running code, and using enterprise tools without requiring new platforms or data formats.

It also includes built-in evaluation and optimisation tools to accelerate performance improvements. Each agent has its own identity, permissions, and guardrails, enabling companies to deploy them in regulated environments.

“It uses Codex to power agents built by companies, third parties, or OpenAI, and makes it easy to securely manage which agents get access to what,” the company’s CEO wrote in a post on X.

Who is using it?

Frontier is already being adopted by major companies including HP, Intuit, Oracle, State Farm, Thermo Fisher, and Uber. Pilot programmes are underway at firms such as BBVA, Cisco, and T-Mobile.

OpenAI is also working with an ecosystem of partners building compatible agent applications using open standards. “We’re working with a small group of Frontier partners—@AbridgeHQ, @AmbienceAI, @clay, @DecagonAI, @harvey, and @SierraPlatform—to design enterprise solutions and support deployment,” the company said in a post on X.


Broader availability is expected in the coming months.