On October 22, 2026, Uber began rolling OpenAI’s ChatGPT Work agent out to more than 30,000 employees, running through ChatGPT Enterprise across its global workforce. It is one of the largest publicly named companywide AI agent deployments so far. For workers, the open question is less whether the tool will show up on their screens and more what it will be allowed to do once it is there.

How Uber’s rollout grows out of a 1,000-person pilot

Uber is not starting from scratch. The company ran a pilot of about 1,000 employees earlier this year, and the new deployment expands that pilot to the whole company. TechCrunch describes the move as Uber’s first companywide deployment of an agentic platform for knowledge work. The sources do not give a precise start date for the pilot, so the internal timeline remains unclear.

The partnership goes back further. Uber and OpenAI first publicly highlighted their work together on May 6, 2026, with an initial focus on AI-assisted tools for drivers and riders. The companywide rollout turns that relationship inward, toward the staff who run Uber’s operations rather than the people who open its app. Crypto Briefing also reported that the deal serves OpenAI as a large reference customer while other big employers weigh similar moves.

What Uber ChatGPT Work actually does

ChatGPT Work is OpenAI’s enterprise agent, introduced in July 2026. It is built for multi-step jobs such as research, data analysis, presentations and workflows, according to OpenAI’s announcement. That is a different product from an ordinary chat window. A chat assistant answers the question in front of it, while an agent takes a goal and works through the steps needed to reach it.

Earlier enterprise chat rollouts mostly gave staff a place to ask questions. An agent that carries out multi-step work touches the files, tools and systems behind those questions. That is why the data-access question matters more here than it did in earlier software rollouts.

Uber is not handing the tool over without structure. Teams in operations, finance, marketing and product are getting access, backed by an internal AI skills library, approved integrations and employee training. unite.ai reports that GPT-6.1 Sol will be the default model inside ChatGPT Work.

Sumanth Sukumar, Uber’s vice president of engineering, said about 80% of early pilot users were non-engineers, according to Traders Union. That figure shows the pilot reached well beyond the engineering organization. Uber CTO Praveen Neppalli Naga framed the goal in the announcement: “As AI transforms how we work, we want to give everyone the tools to innovate, solve problems and help shape Uber’s future.” OpenAI president and co-founder Greg Brockman said Uber is “putting AI to work across its business.”

Where the job question stands after the 3,300-role cut

Uber cut about 3,300 roles in September 2026 as part of an efficiency push that has included growing internal AI use, according to Crypto Briefing. The ChatGPT Work rollout was announced in October, after that cut. The sources do not link the two directly, and Uber has not said ChatGPT Work replaced any specific roles. Readers will likely connect them anyway, but the public record in the sources does not establish cause.

CEO Dara Khosrowshahi has said AI use is becoming a necessity inside the company rather than an optional extra. That points toward more AI-driven work. It does not answer whether headcount will shrink further or which functions will change most.

For scale, Uber reported about 30,400 employees as of December 31, 2023, in its 10-K filing with the SEC. The “more than 30,000” access figure therefore covers roughly the company’s size at the end of 2023. The sources do not include a current headcount after the September cut, so treat that comparison as a snapshot rather than a census.

Labor disputes over AI gains are not confined to software. USA 24 reported that Micron’s Taiwan workers struck over how AI profits are shared. Companies that expand AI tools to large staffs will face similar questions about who captures the productivity gains.

What Uber has not disclosed about cost and results

Uber did not disclose what the deployment costs, according to a Newscord roundup of the coverage. The reporting gives no seat price, contract value or total spend. Anyone trying to benchmark the deal against their own budget has no public number to start from.

Productivity is the other gap. Uber says it will monitor usage and the productivity effect, but it has released no results. Until it does, the case for the rollout rests on the pilot’s reception and executives’ statements, not on measured output.

Most of the detail in this story comes from OpenAI’s announcement and trade press. The coverage reviewed did not include a post written by Uber about the rollout, which means the company’s own framing is still limited to what it has shared through partners and executives.

Practical steps for workers and managers

Workers can start with a few direct questions for their employer or manager:

  • Is an approved AI agent available to me, and which one?
  • What company data can it reach, including files, email and internal tools?
  • What training is provided, and which integrations have been approved?

Managers should settle those answers before a tool goes live, not after. Uber’s rollout pairs the agent with a skills library, approved integrations and training, which gives other employers a usable checklist. Agents that run multi-step tasks also raise the question of who reviews the output, and that answer should be written down.

Business owners weighing a similar move should compare seat pricing and governance policies before signing anything. Because Uber’s price is not public, buyers will need to collect their own vendor quotes. Government buyers already face their own AI purchasing rules, which USA 24 covered in its story on how GSA clause 552.239-7001 shapes the way agencies buy LLMs.

What to watch next as ChatGPT Work spreads

Three things will show whether this rollout is a template or an outlier. First, whether Uber publishes usage or productivity data, which it has said it will monitor but has not shared. Second, whether the company discloses cost figures or changes its headcount plans. Third, whether other large US employers announce similar companywide deployments.

OpenAI has a stake in the answer. The company is nearing a raise of about $30 billion at a $1.25 trillion valuation, led by UAE investors, as USA 24 reported. Large enterprise customers like Uber give investors evidence that the agent business extends well beyond consumer products.

The next test is whether Uber can show measured gains from a 30,000-person rollout while the job questions raised by the September cut remain open. The company has the scale to make that case. It has not yet published the numbers.