Zeta Global and Snowflake said the Open Standard for Interoperability (OSI) could standardize marketing data for enterprise AI, addressing inconsistent data definitions across fragmented systems. Zeta CTO Christian Monberg stated that AI systems often produce inaccurate results when key marketing metrics are defined differently across tools like CDPs, campaign platforms, and attribution systems, noting that "the most common failure mode we see in enterprise AI traces back to the data layer." Monberg added that OSI could standardize definitions, reduce integration work, and improve trust in AI outputs, while future AI marketing will depend on "plain language" queries requiring consistent definitions. Snowflake's Josh Klahr, director of analytics product management, said OSI is gaining early traction with "nearly 400 GitHub stars, more than 40 pull requests, and participation from 48 leading organizations," including Databricks, Salesforce, and Oracle. Klahr expects adoption to grow over the next 12–24 months, with strongest interest in financial services, marketing, and retail. OSI includes five working groups focused on metrics, composability, catalog integration, ontology representation, and semantic tooling, with a community roadmap built from 70+ discussions outlining future work on semantic governance and AI-native systems.