Uber deploys ChatGPT Work to 30,000 employees, expanding agentic AI across enterprise workflows
Uber expands ChatGPT Work from a 1,000-person pilot to its entire global workforce, integrating OpenAI's frontier models with existing enterprise tools.
Last verified:
According to the OpenAI Blog, Uber scaled ChatGPT Work from an initial 1,000-person pilot to its full 30,000-employee global workforce on October 22, marking the company’s first enterprise-wide deployment of an agentic AI platform. The expansion grants access to a multi-tool AI agent designed to orchestrate workflows across Gmail, Google Drive, Slack, and Google Calendar—systems that already anchor Uber’s knowledge work infrastructure.
ChatGPT Work’s integration architecture
The agent’s utility hinges on deep connectivity to Uber’s existing application stack. According to OpenAI’s announcement, ChatGPT Work enables employees to synthesize information across systems—researching, preparing meeting briefs, generating presentations, and completing data analysis tasks without context-switching between applications. Rather than returning isolated suggestions, the platform chains actions across tools, reducing manual handoffs that typically fragment knowledge work.
Model selection and cost-performance tradeoff
Uber will deploy GPT-6.1 Sol as the default reasoning engine within ChatGPT Work. According to the blog, this model selection reflects a deliberate balance between frontier performance and cost efficiency—a critical constraint when scaling reasoning models across tens of thousands of concurrent users. The platform will also surface OpenAI’s multimodal capabilities, including voice and image generation, allowing teams to select the appropriate model variant for specific task types.
Rollout scope and organizational support
The deployment spans multiple business functions. According to the announcement, operations, finance, marketing, and product teams will have access, supported by Uber’s internal AI skills library, pre-approved integrations, and structured training programs. Software engineering teams will continue using Codex, OpenAI’s code-generation tool, to accelerate development workflows—effectively running dual agentic systems optimized for different knowledge domains.
Why This Matters
This rollout signals a fundamental shift in enterprise AI deployment: from isolated chatbots augmenting individual workers to autonomous agents orchestrating cross-application workflows. For enterprises evaluating AI adoption strategies, Uber’s scale demonstrates that agentic systems can operate at production volume without requiring architectural overhauls—they layer onto existing tool stacks. The GPT-6.1 Sol selection also sets a precedent for cost-conscious deployment of frontier models; if the model performs adequately in production, it becomes a reference point for other enterprises weighing the frontier-vs.-efficiency tradeoff. OpenAI’s commitment to track adoption metrics and publish outcomes suggests the company views Uber as both a validation of ChatGPT Work’s viability and a template for future enterprise rollouts.
Frequently Asked Questions
What is ChatGPT Work and what can it do?
ChatGPT Work is an agentic AI platform within ChatGPT Enterprise that handles multi-step tasks across integrated applications. It can research information, create presentations, analyze data, prepare meeting briefs, and execute workflows by connecting to Gmail, Google Drive, Slack, and Google Calendar.
Which model powers ChatGPT Work at Uber?
GPT-6.1 Sol serves as the default model for Uber's ChatGPT Work deployment. According to OpenAI, this model balances frontier capabilities with cost efficiency for enterprise use.
Why is this rollout significant for enterprise AI adoption?
This represents a major shift from AI copilots that augment individual tasks to agentic systems that autonomously complete workflows across multiple tools, demonstrating multi-application reasoning at scale across 30,000 workers.