OpenAI adds a Data agent for conversational business analysis

OpenAI is introducing a Data agent in ChatGPT Work that lets employees investigate company data through natural-language conversations. The product announcement, published by OpenAI News on September 10, 2026, positions the tool as a way to get answers and build dashboards without writing queries or learning a separate analytics platform.

The agent can connect to approved sources including Amazon Redshift, Datadog, Google BigQuery, ClickHouse, Databricks, MongoDB, and Snowflake. It can also use files and documents stored in Google Drive and SharePoint. Rather than treating data as a collection of isolated tables, it can use business definitions, relationships, custom calculations, and semantic context from systems such as dbt, GitHub, Snowflake Horizon, BI dashboards, and Databricks Genie Ontology.

Why it matters for AI builders

Users can ask follow-up questions, inspect the evidence behind findings, and convert an analysis into an interactive dashboard with visualizations. Dashboards can be edited, shared, refreshed, and styled with an organization’s brand guidelines. OpenAI’s examples include tracking customer adoption, identifying drop-off points, and monitoring activity across products and subscription tiers.

For teams building AI workflows, governance is a central detail. Administrators decide which connections and roles are available, while queries inherit the connected account’s existing table-, row-, and column-level permissions. That could make conversational analytics easier to deploy in organizations that already have established access controls, though the usefulness of results will still depend on the quality and clarity of the underlying business data.

Source: OpenAI News