OpenAI has introduced a new Data agent in ChatGPT Work, an agent the company says connects to a company's approved data sources, investigates why business metrics changed, and builds interactive dashboards that employees can share — all through a plain-language conversation, without writing SQL queries or learning a separate analytics tool.
The launch, announced Thursday, is OpenAI's most direct move yet into enterprise business intelligence, a market where Snowflake, Databricks, Microsoft and a wave of analytics startups are racing to add AI analysts of their own. According to Unite.AI's coverage of the announcement, the agent is framed around everyday business questions — why sales slowed last quarter, where spending is rising, which accounts are at risk of churning — that today often mean waiting days for a report or pulling a data team into a meeting. Anyone following breaking AI news this year will recognize the pattern: the vendors that own the workflow, not just the model, win the budget.
How the Data Agent Works
Users direct and refine an analysis in a single conversation. The agent can investigate what changed in a business metric, trace contributing factors across connected sources, and present its findings with citations. OpenAI demonstrated the flow with a fictional "Acme Cloud" example, where a single prompt produced an adoption dashboard covering June 2 to August 21, 2026, with sections for installs, activation, activity and retention — generated by asking the agent to track adoption from first install through ongoing usage and flag where customers get stuck.
Beyond analysis, OpenAI says ChatGPT Work can recommend next steps, identify who needs to be involved, share findings through Slack or email, and carry out user-approved actions through connected tools.
Data Connections and Governance
The Data agent connects to a long list of enterprise data platforms, including Amazon Redshift, Datadog, Google BigQuery, ClickHouse, Databricks, MongoDB and Snowflake, and can pull files and documents from Google Drive and SharePoint into an analysis.
Crucially for regulated industries, the agent is designed to interpret data the way the organization itself does. OpenAI says it applies a company's own business terms, metric definitions, custom calculations and data relationships, drawing context from semantic layers and trusted sources such as Databricks Genie Ontology, dbt, GitHub, Snowflake Horizon and existing business-intelligence dashboards.
Governance is enforced at the source. Enterprise administrators choose which data connections exist and which roles can use them, and queries run under the connected account's existing permissions, including table-, row- and column-level restrictions. The company pairs that role-based access control with citations and reviews for generated outputs.
Dashboards That Live in Existing BI Tools
An analysis can be turned into an interactive dashboard with built-in visualizations that teammates can edit, share and refresh, and companies can supply brand guidelines so outputs match their look and feel.
OpenAI also says the agent can build and interact with dashboards inside existing business-intelligence platforms — Omni, Oracle BI, Power BI, Sigma, Tableau and ThoughtSpot. Bogdan Crivat, corporate vice president of Microsoft Fabric, said in the announcement that users can create Power BI dashboards simply by describing their business questions.
Early Adoption Inside and Outside OpenAI
OpenAI claims significant internal usage: nearly all of its product team and more than two-thirds of its go-to-market organization already use data agents in ChatGPT Work. The company credits an internal data team that created shared business definitions, set access rules and put safeguards around sensitive data.
Externally, OpenAI says organizations including NTT Data, Thermo Fisher and ServiceTitan are using the agent through its Alpha program.
The announcement carried partner statements from AWS, ClickHouse, Databricks, Snowflake, MongoDB and G2. Pablo Stern, MongoDB's chief product officer for AI and emerging products, said the agent can reach data stored in Atlas as it updates in real time. Umesh Unnikrishnan, Snowflake's head of developer experiences, said employees can draw on Snowflake data under their existing access controls.
A Crowded, Benchmark-Free Race
The launch lands in a market moving fast. Databricks, Snowflake and Microsoft have all shipped agentic analytics features of their own, and startups are competing for the same "AI analyst" positioning. VentureBeat noted that OpenAI skipped the one thing several rivals are racing to publish: a public benchmark demonstrating the agent's accuracy. That absence could draw scrutiny, particularly given how much weight enterprise buyers place on reliability figures when numbers drive decisions.
Coverage from Seeking Alpha also highlighted that the release was accompanied by a new Agents API, extending OpenAI's push from a single chat product toward a platform of composable agents for enterprise workflows.
For OpenAI, the bet is that the natural-language interface becomes the front door to enterprise data, with the underlying warehouse reduced to plumbing. For incumbents, the counter-strategy is visible in the launch itself: OpenAI had to integrate with all of them to make its agent useful at all. Which layer captures the value — the assistant or the data platform beneath it — is now the central question in enterprise AI.
