Organizations Have Restructured Data for AI Analysis
New database architectures now enable AI systems to retrieve and interpret statistics without human intervention.
Updated on Oct. 6, 2026 in Artificial Intelligence
Live Poll
Do you trust AI-driven automation to handle complex data tasks accurately in your workplace?
Major institutions including Bloomberg, the UN, and Snowflake have overhauled their data architectures to improve AI accessibility. These systems now allow AI to directly query and analyze statistical datasets using everyday language.
Why it matters
Traditional data silos required human analysts to manually verify metrics and identifiers, creating a bottleneck for AI processing. Restructuring these databases ensures that statistical definitions are machine-readable, reducing the risk of errors in AI-driven analysis.
Snowflake's AI service handled over 330,000 inquiries by the end of 2025, while Siemens now processes 2,800 customer inquiries weekly through automated systems. These figures reflect a shift from manual database lookups to direct API-based statistical retrieval.
The players
Bloomberg
A financial data and media company providing market infrastructure and analytics.
Snowflake
A cloud-based data warehousing company offering specialized AI services for enterprise data analysis.
Siemens
A German industrial manufacturing company integrating AI for customer service and internal operations.
United Nations
An international organization managing the UN System Data Commons to standardize global development data.
The details
Organizations are modifying their data architectures to bundle statistical definitions—formal descriptions of how a metric is calculated—alongside numerical values. This allows AI to access metadata directly from institutional platforms like the UN's Data360, rather than relying on human verification to define parameters. By connecting AI directly to these structured environments, systems can parse complex queries like South Korea's 2025 employment rate of 63.1% without human oversight.
Timeline
By the end of 2025, Snowflake's AI service processed 330,000 inquiries.
September 17, 2026: The UN unveiled the UN System Data Commons platform.
September 29, 2026: Bloomberg announced its new data provision method.
2027: The UN targets 80% of its total statistical data to be linked for AI access.
The Tech Race
This movement follows the trajectory set by the UN System Data Commons platform, which aims to centralize fragmented global data. It marks a departure from proprietary, siloed databases toward an interoperable standard designed specifically for AI consumption.
Users will eventually see improved accuracy in financial and scientific reporting as AI assistants gain direct access to verified institutional datasets. While these tools currently serve enterprise employees and researchers, they will soon shift the baseline for how institutional statistics are queried.
The takeaway
The era of manual database searching for AI analysis is ending as institutions standardize data for automated retrieval. Watch the progress toward the UN’s 2027 goal of linking 80% of its data, which will serve as a bellwether for institutional AI readiness.
What happens next
The United Nations aims to link 80% of its total statistical data by 2027.
Further reading
For more on the underlying systems, visit the Artificial Intelligence section.
Source note: This article includes information reported by 조선일보.
Live Poll
Do you trust AI-driven automation to handle complex data tasks accurately in your workplace?
