ChatGPT Work Tested With 30,000 Uber Staff In AI Trial

Uber is rolling out ChatGPT Work to more than 30,000 employees, giving OpenAI one of its largest enterprise deployments as the company embeds AI into research, data analysis and internal workflows.

Key Takeaways

Uber is rolling out ChatGPT Work to more than 30,000 employees on Thursday
OpenAI said Uber will use ChatGPT Work across teams for research, data analysis, presentation building and more complex internal workflows
OpenAI also posted the rollout on X, underscoring that the company sees the Uber deployment as a flagship enterprise case
The rollout tests whether a single assistant can become normal office infrastructure across varied roles like operations, product, support and corporate

OpenAI said Uber will use ChatGPT Work across teams for research, data analysis, presentation building and more complex internal workflows. Uber runs one of the largest global transportation and delivery networks, so a companywide deployment puts AI tools in front of workers across product, operations and corporate functions.

ChatGPT Work Reaches Uber At Scale

The operative number is 30,000-plus employees, not a narrow pilot, which matters because many enterprise AI deals still start with a few hundred users in legal, support or software teams.

A broad internal rollout changes the test by asking whether workers will use one system for many tasks instead of a patchwork of single-purpose tools. In practice, that means fewer separate apps for drafting, summarizing, spreadsheet analysis and slide creation.

OpenAI also posted the rollout on X, underscoring that the company sees the Uber deployment as a flagship enterprise case rather than a quiet customer win.

From Team Pilots To Companywide Use

Large companies spent much of 2025 and early 2026 fencing generative AI into controlled pilots, with legal teams trying contract review, finance teams testing memo drafting, and engineers using code assistants in ring-fenced environments.

This rollout moves past that stage, Uber is no longer testing whether generative AI can help one department, it is deciding that a shared tool belongs across the business, which shifts the main question from capability to governance, access control and return on seat costs.

The timing also fits a wider shift in enterprise buying.

Instead of asking whether a chatbot can draft text, buyers are asking whether a single assistant can tie together search, analysis and action inside existing work.

Also Read: U.S. Says Japan Deal Hands New York 2 Chip Plants And $21B

How Uber’s AI Rollout Changes The Buyer Playbook

ChatGPT Work sits in the layer between a general chatbot and a full custom software build. OpenAI packages a base model with enterprise controls, then adds connectors, memory and task support so employees can use it for real internal work.

For CIOs, that changes procurement math, as a company like Uber can buy one broad platform and let many teams adapt it, instead of licensing separate niche AI tools for each workflow.

That can cut integration work, but it also concentrates risk in one vendor if pricing, uptime or policy terms change.

For employees, the difference is simpler. ChatGPT Work is meant to handle everyday knowledge tasks that used to bounce between search, spreadsheets, docs and slide tools.

If adoption sticks at Uber, other large employers will have a stronger case for moving beyond pilot groups.

What to watch next is usage depth, not just seat count. The headline number tells you how many workers can access the system.

It does not show how many will use it daily, or which tasks save the most time.

What 30,000 Seats Actually Test

A 30,000-seat rollout tests something basic that AI vendors rarely spell out, can one assistant become normal office infrastructure. A pilot proves a model can answer prompts, while a deployment this large tests whether employees trust it with repeated work, whether managers can measure gains, and whether security teams can live with its data paths.

Uber is a useful case because its staff spans operations, product, support and corporate roles.

If ChatGPT Work lands there, the product is not just a writing bot for one white-collar niche.

It becomes a shared layer for varied business tasks, from preparing internal briefings to pulling patterns from large internal datasets. The next milestone is likely not another headline seat number but evidence on retention, workflow depth and whether Uber builds more of its own internal processes around OpenAI’s stack.