Enterprise AI Adoption

OpenAI emphasizes that companies are no longer just testing AI, but have moved into a stage of actually deploying it across enterprise-wide operations. CRO Denise Dresser says that during her first 90 days meeting with hundreds of customers, she confirmed a strong sense of urgency spreading across the industry to redesign companies around AI.

On the business side, enterprise demand is also growing rapidly. Enterprise revenue accounts for more than 40% of the total, and is projected to reach a level comparable to consumer revenue by the end of 2026. At the same time, Codex has 3 million weekly active users, the API processes more than 15 billion tokens per minute, and GPT-5.4 is driving record engagement in agentic workflows.

The company now frames enterprise AI adoption around two core questions:

• How to apply the most powerful AI not just as department-specific tools, but across the entire business
• How to naturally weave AI into employees' everyday workflows

OpenAI's proposed answer is to position Frontier as the intelligence layer overseeing agents across the whole enterprise, and to make the unified AI superapp the primary interface through which employees get work done. OpenAI presents as its strength the fact that it is one of the few companies building the model, the infrastructure, and the interface employees actually use, all together.

A particular pain point for enterprise customers is disconnected AI point solutions. To address this, OpenAI aims to provide a unified operating layer that connects a company's context, internal systems, external data, and permissions and controls, and explains that OpenAI Frontier is helping customers like Oracle, State Farm, and Uber deploy agents company-wide.

OpenAI is also strengthening its strategy of attaching its intelligence to existing infrastructure and data ecosystems through Frontier Alliances partners such as McKinsey, BCG, Accenture, and Capgemini, as well as partners like AWS, Databricks, and Snowflake. The Stateful Runtime Environment, built together with AWS, helps agents maintain context, remember previous work, and carry out more complex real-world tasks across multiple tools and data sources.

A product strategy that bridges personal and work use is also central. OpenAI aims to have employees collaborate with AI throughout the day through a single experience spanning ChatGPT, Codex, and agentic browsing. Indeed, Codex has grown more than 5x since the start of the year, and customers such as GitHub, Nextdoor, Notion, and Wonderful are building multi-agent systems that carry engineering tasks through to completion.

Internally, use cases are also expanding. For example, sales teams use agents to research inbound leads, score them against qualification criteria, send personalized emails, and even update the CRM to drive new business. OpenAI believes that its mass consumer base of 900 million weekly ChatGPT users reduces friction in enterprise adoption and accelerates employees offloading repetitive tasks to agents so they can focus on bigger challenges.