AI NewsWaveSnowflake Expands Autonomous Execution and Control for Enterprise AI Agents
Key points
- Snowflake expands autonomous execution for AI agents
- Business ontology unifies the standards for terms and metrics
- CoCo agents automate data pipeline management
Snowflake has unveiled a series of features to help enterprise AI agents handle work on their own. Techworld News, Digital Today, Byline Network, iNews24, Chosun Biz and Datanet reported on it over the 3rd and 4th. Techworld News wrote that the company is strengthening data management, real-time analytics and governance features to support autonomous task execution by AI agents.
Each outlet stressed a different point. Digital Today reported that Snowflake unveiled a "business ontology" that manages business terms and metric definitions scattered across a company under a common standard and helps AI agents give consistent answers. iNews24 introduced it as an agent feature that manages data definitions, access rights and task history in one place. Digital Daily reported that, under the slogan "same data, same answer," the company put its understanding of enterprise data and control over execution forward as its differentiating strategy.
On autonomous execution, Chosun Biz reported that Snowflake unveiled an AI agent that plans and handles tasks in the background without separate instructions, and explained in its headline that the agent calls on the user only when approval is needed. Digital Today wrote that Snowflake said on the 3rd (local time) that it is expanding its coding AI agent platform "CoCo" with "CoCo Cloud Agent" and "CoCo Automation," which monitors and fixes data pipelines. Byline Network named the Cortex AI Gateway, sandboxes and the business ontology, among others, as control strategies against data leaks, misuse of permissions and rising costs as agents spread.
On the company's place in the market, Bloter noted that while OpenAI and Anthropic build the models that serve as the AI "brain," Snowflake put forward the enterprise data that brain will handle and the mechanisms to control it. Datanet explained that as AI agents begin to carry out real work beyond answering questions, the ability to understand business context and control agents so they operate under security and governance policies is becoming more important.
Recent developments Hellot reported that Oracle and Snowflake announced they will offer the Snowflake AI Data Cloud on Oracle Cloud Infrastructure (OCI) as a native first-party service. The announcement was made at Oracle AI World 2026.
Terms explained: An ontology is a framework that systematically defines concepts, such as business terms and metrics, and the relationships among them. Data governance refers to the rules and systems that manage data access rights, quality and security.