OpenAI Launches 'Data agent' for ChatGPT Work: Enterprise Data and Dashboards

OpenAI has launched 'Data agent' for ChatGPT Work, enabling conversational data analysis across Snowflake, BigQuery, and Databricks alongside interactive dashbo

tau · September 11, 2026

#OpenAI #ChatGPT Work #DataAgent #DataAnalysis #Dashboards #Snowflake #BigQuery #Databricks #AINews

OpenAI Launches 'Data agent' for ChatGPT Work: Enterprise Data and Dashboards

On September 10, 2026, OpenAI officially introduced 'Data agent' for its enterprise ChatGPT Work environment, empowering corporate teams to query distributed company data conversationally and assemble real-time interactive dashboards. The release is engineered to eliminate traditional analytics bottlenecks, allowing business stakeholders to investigate operational anomalies and uncover root causes through natural language dialogue without writing custom SQL queries or mastering dedicated business intelligence suites.

OpenAI 'Now everyone can put data to work' - Official announcement graphic for ChatGPT Work Data agent

Image source: OpenAI

Direct Integration Across Data Warehouses and Enterprise Repositories

The architectural core of Data agent centers on unifying fragmented enterprise infrastructure directly within a single conversational interface.

The agent connects directly to an extensive portfolio of sanctioned databases and cloud data warehouses. Supported structured data sources include Amazon Redshift, Datadog, Google BigQuery, ClickHouse, Databricks, MongoDB, and Snowflake, providing robust operational coverage for high-throughput transactional records and large-scale time-series telemetry.

Beyond structured data engines, Data agent bridges unstructured organizational knowledge by integrating Google Drive and Microsoft SharePoint repositories. By synthesizing numeric warehouse entries alongside internal proposals, meeting logs, and slide decks in a shared analytical context, business teams can evaluate quantitative variance against strategic narrative documentation.

Plugins Interface, '@Data' Mention Workflow, and Dynamic Dashboards

From an end-user standpoint, Data agent operates seamlessly within the ChatGPT Work workspace application ecosystem.

Once enterprise administrators configure and enable the 'Data' plugin from the workspace Plugins directory, employees can invoke the capability inside standard conversation threads simply by typing the @Data mention. Users can ask conversational questions such as identifying sudden regional revenue shifts or auditing unexpected traffic spikes, prompting the agent to fetch relevant records from Snowflake or BigQuery and conduct step-by-step root cause analysis.

Rather than returning plain text summaries, the agent renders interactive dashboards natively within the chat stream. Decision-makers can filter categorical variables, adjust temporal windows dynamically, and share dashboard URLs across functional teams for immediate cross-departmental coordination.

Workspace Rollout Requirements and Enterprise Governance Discipline

Organizations evaluating enterprise adoption must account for clear infrastructural prerequisites and administrative governance responsibilities.

Data agent is exclusively deployed within the enterprise-focused 'ChatGPT Work' tier through its Plugins directory and is not accessible to personal Free or ChatGPT Plus tiers. To protect confidential company information, workspace administrators must formally activate the plugin and grant vetted connection credentials for individual platforms such as Databricks and Snowflake before end users can initialize queries.

Furthermore, engineering and security leaders must evaluate write and action permissions on connected systems alongside read-only telemetry. Maintaining rigorous audit trails, data reconciliation standards, and compliance boundaries remains an essential prerequisite for organizations deploying autonomous data agents across production environments.

Sources