Uber brings ChatGPT Work to all employees with OpenAI
- ID
- 38346
- Status
- summarized
- Published
- 22 Oct 2026, 8:00 AM
- Fetched
- 23 Oct 2026, 1:21 AM
- Provider
- OpenAI News
- Category
- ai-labs
- Original URL
- https://openai.com/index/uber-chatgpt-work-rollout
- Source URL
- https://openai.com/news/rss.xml
Summary
- Score
- 4.0
- Created
- 23 Oct 2026, 1:21 AM
- Tags
- Audience
- ai_agent_userssaas_foundersai_ml_learners
What happened
OpenAI's post says Uber is expanding ChatGPT Work — an agent inside ChatGPT Enterprise that connects to Google Drive, Gmail, Google Calendar and Slack — from a 1,000-employee pilot to more than 30,000 employees company-wide, described as Uber's first company-wide agentic platform deployment for knowledge work. GPT-6.1 Sol is named as the default model inside ChatGPT Work, positioned as balancing frontier performance and cost efficiency, with multimodal voice and image capabilities also available. Uber's engineering teams continue using Codex separately, and the rollout is backed by an internal AI skills library, approved integrations and employee training.
Why it matters
This is a vendor-written customer story, so treat it as a signal about packaging rather than proof of results: the notable detail is that a 30,000-seat agent rollout defaults to a single cost-efficiency-tiered model (GPT-6.1 Sol) rather than the top model, which is the same tradeoff you face when choosing a default model for your own agents. If you're building or buying internal agent tooling, the listed integration surface — Drive, Gmail, Calendar, Slack — plus a curated 'approved integrations' list and a skills library is a concrete checklist for what a deployable internal agent needs beyond the model itself.
Discussion angle
Would you default your internal agent to a cheaper 'cost efficiency' model tier like Uber's GPT-6.1 Sol choice, or route per task — and what breaks when 30,000 non-engineers get an agent with write access to Drive, Gmail and Slack?