# Why trusted context is becoming the currency for enterprise AI > Source: > Published: 2026-07-07 09:00:00+00:00 AI is getting most of the attention in enterprise technology. Governance, ownership, and data quality do most of the heavy lifting behind the scenes. And yet, as organizations move from AI experiments to production deployments, trusted context is becoming a key factor in determining whether agents create business value — or operational risk. That shift is reshaping how Salesforce, Microsoft, Snowflake, Databricks, SAP, Oracle, and others are positioning their data, governance, metadata, and integration services. The conversation is no longer just about models. It’s about whether AI systems can operate against trusted, governed, and business-relevant information. Trusted context has become the new currency, and Salesforce has made a strategic commitment to it. Master data management (MDM) spent much of the last decade as an important but often overlooked infrastructure. AI is changing that. Agentic systems can identify duplicate records, inconsistent definitions, fragmented ownership, and poor governance the moment AI begins interacting with enterprise data and processes. I recently wrote about [Salesforce’s State of Data and Analytics research](https://www.forbes.com/sites/moorinsights/2026/01/15/weak-data-management-hinders-enterprise-ai-salesforce-research-shows/), which found that 84% of data leaders believe their organizations need significant changes to their data strategies before AI can succeed at scale. That finding shows what many enterprises are now experiencing. AI often exposes data and governance issues that have existed for years. [Manouj Tahiliani](https://www.linkedin.com/in/manoujtahiliani/), senior vice president for MDM at Informatica, now part of Salesforce, said, “Trusted context is becoming the new currency in enterprise AI.” His argument is that trusted context is the connected, governed view of customers, products, and suppliers that lets an agent act like a tenured employee. Models and agents will commoditize. Differentiation comes from how well an agent understands the enterprise, which depends on the data underneath. AI is not a model problem. It is a data foundation problem with an agent interface bolted on top. Salesforce completed its acquisition of Informatica in November 2025. The acquisition strengthens Salesforce’s position around data quality, governance, metadata, lineage, and MDM. It also reflects the market reality. Every major enterprise platform provider is trying to create a trusted layer that connects operational systems, business context, and AI. Marc Benioff, CEO of Salesforce, summarized the rationale when the deal closed. Organizations need trusted, connected, and governed data before they can expect meaningful outcomes from AI. While that statement may sound obvious, it reflects one of the biggest challenges organizations continue to face as AI moves into production. The combined strategy brings together Tableau for analytics, MuleSoft for integration and Agent Fabric, Data 360 (formerly Data Cloud) for data unification, and Informatica for governance, quality, stewardship, and MDM. The goal is not simply data consolidation. The goal is creating a consistent layer of business context that can be used across applications, workflows, and AI systems. Salesforce is not alone. Microsoft, for example, is building around Fabric, OneLake, Purview, and Fabric IQ. Snowflake continues expanding governance, semantic, and catalog capabilities. Databricks is advancing Unity Catalog and its broader Data Intelligence Platform strategy. SAP and Oracle are pursuing similar objectives through business applications and industry-specific data models. The competitive landscape is increasingly shifting from data storage and analytics toward trusted context, governance, and operational execution. Early adoption metrics suggest the strategy is gaining traction, although long-term success will be measured by customer outcomes, implementation timelines, and operational value. Data 360 has grown within Salesforce, Agentforce adoption continues to expand, and deeper integration between Informatica, Data 360, and Agent Fabric is expected throughout 2026. The Intelligent Data Management Cloud (IDMC) remains the foundation underneath Informatica’s data management strategy. It provides metadata-aware connectivity, governance, stewardship, matching, merging, and master data capabilities across applications, databases, files, and streaming sources. For most enterprises, the number of connectors is less important than whether governance, ownership, quality, and lineage remain consistent across systems. Connectivity alone rarely solves data problems. Operational discipline does. What is changing is how those capabilities are being exposed to AI systems. Salesforce and Informatica are positioning governance and data management services as capabilities that agents can access directly through [Model Context Protocol](https://www.infoworld.com/article/4029634/what-is-model-context-protocol-how-mcp-bridges-ai-and-external-services.html) and related interfaces. The value is not the protocol itself. The value is allowing AI systems to interact with governed enterprise information while maintaining lineage, governance, ownership, and security controls. Headless data management is also becoming more important. Organizations want agents, applications, and workflows to access trusted services without custom integrations for every use case. If executed effectively, that approach could simplify how AI systems consume enterprise data while preserving governance standards. Industry research has consistently shown that many MDM initiatives struggle to achieve their original business objectives. Governance arrives too late. Executive sponsorship is weak. Ownership remains unclear. Business units maintain competing definitions. Technology is expected to solve organizational problems. One of the recurring issues I see across enterprises is that technology decisions often move faster than governance models. Organizations frequently deploy tools before establishing ownership, stewardship, and accountability. AI tends to expose those gaps very quickly. The challenge becomes more complicated as enterprises deploy agents across ERP, CRM, finance, supply chain, and operational systems simultaneously. Visibility, accountability, and governance become increasingly important as AI systems move beyond recommendations and begin to influence business processes. This is where Informatica’s Agent Fabric Context Catalog becomes relevant. The concept is less about cataloging technology and more about providing visibility into how agents are deployed, governed, monitored, and controlled. Tahiliani offered advice that aligns with what I often tell clients. Start with business priorities. Translate those priorities into a data strategy. Then select the architecture and technology required to support it. Many organizations still approach the process in reverse, struggling to generate business value. MDM is not a single-vendor market. Gartner’s 2026 Magic Quadrant leaders include Salesforce (Informatica), Profisee, Reltio, Semarchy, and Stibo Systems. Each vendor approaches the market differently. Profisee remains closely aligned with Microsoft environments. Reltio, which SAP acquired in May 2026, continues to differentiate through graph-oriented architecture and API-first design. Semarchy brings strengths where integration and MDM converge. Stibo maintains a strong position in product information management and retail-focused environments. Informatica’s key strengths continue to be its broad capabilities, mature governance, and growing alignment with Salesforce. The larger question is execution. Enterprises will want evidence that implementation timelines, governance complexity, and time-to-value improve as the roadmap evolves. Historically, Informatica implementations have required significant investment, governance discipline, and organizational commitment. Salesforce will need to demonstrate that the combined strategy can simplify adoption while maintaining the governance rigor many customers expect. Yum Brands, the parent company of KFC, Pizza Hut, Taco Bell, and Habit Burger Grill, operates more than 63,000 restaurant locations globally. According to company leadership, significant effort was being spent consolidating and cleansing location data before it could be used effectively across the business. Informatica MDM became a central component of the company’s modernization effort. TELUS represents a different use case. The Canadian telecommunications and health services provider uses Informatica MDM Cloud Edition and Customer 360 to improve customer visibility across the organization. Integrating acquisition data into a unified customer view enabled more effective measurement of marketing performance and improved opportunities for targeted cross-sell initiatives. Neither example proves the broader strategy on its own. Both illustrate a pattern that continues to emerge across enterprise AI initiatives. Data management investments create value when they improve operational execution, decision-making, and business outcomes rather than simply improving data quality metrics. The common theme is that trusted information is becoming a foundational requirement for organizations attempting to scale AI, analytics, and operational decision-making. The questions that separate successful data programs from costly tech projects are straightforward. Is there clear ownership for each data domain? Is governance embedded from the beginning rather than added later? Can governance and data management services be consumed directly by AI systems? Can compliance, security, and operational controls scale alongside AI adoption? These questions matter more than any individual AI feature announcement. For Salesforce, the next phase requires measurable proof points. Customer references are encouraging, but enterprises will want audited outcomes, implementation metrics, and long-term operational results. I believe that success in enterprise AI won’t come from having the best model. Instead, it will come from the team with the clearest, best-governed data to support their efforts. This reflects how ERP systems are evolving, not being replaced, with an emphasis on enhancing the core data rather than just updating the technology. Salesforce has made a decisive commitment to making trusted context essential to enterprise AI, setting a high standard that all other vendors must meet. The proof will not be in the keynotes. It will be in the stores Yum can finally report on, the households TELUS can finally sell into, and the next 10 customer stories about successful AI integration. — *Disclosure:** KramerERP offers paid services to technology companies, similar to those provided by other technology research and analyst firms. These services include research, analysis, advisory services, consulting, benchmarking, acquisition matchmaking, video sponsorships, speaking sponsorships and other related activities. KramerERP has worked with, or is currently working with, companies mentioned in this article.*