At a recent media roundtable, leaders from IBM, Salesforce, and Snowflake highlighted how Zero Copy technology is dismantling data silos to fuel AI and real-time business applications.

Why it matters: As enterprises race to adopt AI, they’re encountering significant roadblocks, including fragmented data and slow integration. Salesforce research shows that security challenges and interoperability issues are in fact the biggest concerns for open ecosystem adoption. 

Driving the news: The panel discussion revealed how three industry leaders are converging on Zero Copy architectures and open standards to deliver immediate, contextual, and governed data for AI use cases that unlock new levels of business value and responsibility.

What they’re saying: 

• Data Governance Should Never Be An Afterthought. IBM’s Vice President, Product Management, watsonx Platform Ed Calvesbert emphasized that data management and governance must be “front and center in the design of these initiatives.” He highlighted the challenge of unifying disparate governance models for structured and unstructured data, stressing the need for flexible enforcement at the point of query and strong vendor collaboration on interoperable solutions.
• Real-Time Context and Cross-Application Interoperability Are Non-Negotiable. Snowflake’s Senior Director of Product Management Sap Mukherjee stressed that AI solutions demand not just raw data, but deep business context and real-time access to enterprise events for agents to act swiftly. He underscored that agentic AI will span across application boundaries, necessitating open standards like Apache Iceberg™ and a control plane (like MCP) for intent routing, alongside a data plane (Zero Copy) for consistent context.
• Unified, Real-Time Data Activation Is Key to Enterprise Transformation. Salesforce Vice President, Product Management Narinder Singh highlighted that the core challenge is breaking down data silos and achieving “data fluidity” to bring trapped data directly to line-of-business users and processes. He emphasized that Salesforce Data Cloud‘s metadata-driven, Zero Copy foundation, combined with Agentforce and Customer 360 applications, addresses latency, enables direct activation, and embeds business-contextual governance.

Open Image Modal

Image Modal

Between the lines: The future of AI depends on a trusted, integrated data layer, not just advanced models. 

• Zero Copy architecture lets businesses use their current data without creating new silos or risking security.
• IBM, Salesforce, Snowflake, and other industry leaders are collaborating to promote open standards and interoperability. This means companies can pick the best tools while keeping data unified and speeding up AI projects.
• The focus is shifting. It’s no longer just about having data — it’s about having actionable, ethically governed data available instantly for autonomous AI.

What’s next: The industry is moving toward a future where data platforms prioritize openness, interoperability, and integrated governance as core tenets. Salesforce experts like Singh expect continued advancements in open data formats (like Apache Iceberg), shared metadata catalogs (like Apache Polaris™), and cross-platform policy enforcement to support the escalating demands of agent-to-agent communication and truly autonomous business processes. Together with a network of partners, Zero Copy architecture delivers unprecedented access to mission-critical enterprise data without moving or copying it with direct data flow between IBM Z systems, mainframes, and structured databases like IBM Db2. 

Go deeper:

• Learn more about why Zero Copy Is Critical to a CIO’s Agentic AI Strategy
• Read more about Building Trust at Scale: Why Enterprise Knowledge Is the Foundation of Trustworthy AI
• Uncover the value of Data Cloud and its unprecedented growth as the intelligent activation engine of the Salesforce Platform