Snowflake Horizon Catalog: Governance & Security for Enterprise AI

New innovations across Snowflake Horizon Catalog centralize AI governance, context, and security to provide a trusted foundation for enterprise AI across data, tools, and agents

• Built for the Agentic Era: New enhancements across Snowflake Horizon Catalog ensure that every person, tool, and agent shares the same trusted business context through richer semantic views, built-in governance, and security controls, with customers including BlackRock using the new Horizon Context to ensure AI operates on a shared definition of enterprise truth

• Proactive Built-in Security for AI: Acxiom, NewDay, and Thomson Reuters are working with Snowflake to explore how Snowflake’s new AI security capabilities can help strengthen security, governance, and control for enterprise AI and agentic systems

• Scale with Simplicity: Working seamlessly with Snowflake Horizon Catalog, enterprises can now scale their AI apps and agents with unrivaled performance and operational simplicity using Adaptive Compute

SAN FRANCISCO, SNOWFLAKE SUMMIT 26 – June 2, 2026 – Snowflake (NYSE: SNOW), the AI Data Cloud company, today announced at Snowflake Summit 26 new innovations across Snowflake Horizon Catalog that redefine how enterprises govern, contextualize, and secure AI.

As organizations move from AI experimentation to autonomous systems operating at enterprise scale, they need AI that operates using trusted business context, while remaining secure, governed, and compliant. Snowflake Horizon Catalog serves as the universal AI catalog for enterprise data. New capabilities like Horizon Context ensure that every person, tool, and AI agent operates from the same trusted business context. Combined with new security innovations that provide purpose-built controls to govern and secure AI agents, Snowflake Horizon Catalog is the connected foundation for trusted AI. Built on the governance, security, and control provided by Horizon Catalog, Adaptive Compute² automatically optimizes compute and software resources in real-time for customers, delivering fast, efficient AI and app performance at enterprise scale, without manual tuning or infrastructure management.

“When intelligence becomes autonomous, trust is no longer an afterthought, it becomes foundational,” said Christian Kleinerman, EVP of Product, Snowflake. “Organizations need AI that operates from trusted business context with governance and security built in from the start. New advancements across Snowflake Horizon Catalog give every agent, app, and team the trusted context and security controls needed to move AI from experimentation into real-world business operations.”

Create a Shared Understanding of Business Data for AI

As AI agents make more decisions on their own, even small inconsistencies in data can lead to critical mistakes. This is why many AI projects stall between proof of concept and production. Traditional semantic layers are separate from the data itself, which makes it hard to maintain consistent definitions and governance across systems. For example, an AI agent might recommend increasing prices based on revenue data, but if revenue is defined or calculated in different ways across systems, that recommendation can lead to misinformed decisions.

Horizon Context solves this by providing the context layer for AI and BI so data has the same meaning everywhere and AI-driven decisions are reliable. Enterprises like BlackRock are already using Horizon Context to ensure AI operates on a shared definition of enterprise truth. Horizon Context brings this to life by enabling enterprises to:

• Collect Trusted Business Context: Today, business logic is fragmented across SQL, BI dashboards, and agents, leading to unreliable AI insights. Horizon Context brings together business context across an organization’s entire data estate, including databases, data lakes, and BI tools, to ensure every tool, team, and AI agent draws from the same trusted context. This enables teams to quickly find, organize, and trust the right data.

• Automatically Maintain Business Context: When LLMs are grounded in business logic they stop guessing and start reasoning accurately. Horizon Context now gives data analysts a richer, connected view across their data by bringing together business context around how data is used, trusted, and kept up to date. Capabilities like Semantic Studio⁴ enable teams to define shared business logic without requiring SQL expertise, while Semantic View Autopilot automatically creates and refines semantic views that maintain this context over time. Semantic Views and data agents can also be automatically created for data shared across organizations, including datasets from Snowflake Marketplace, carrying trusted business context into every AI and analytics workflow. By giving AI access to consistent, governed business definitions, teams can simply ask a business question to Snowflake CoCo, the coding agent where you build faster, and get a trusted answer, even if the data is outside of Snowflake.

• Extend Trusted Context Everywhere: Snowflake is expanding access to trusted business context across the enterprise, powering more reliable AI and analytics. With Horizon Context, CoCo delivers more trustworthy and consistent responses by automatically pulling in rich business context grounded in trusted enterprise knowledge. Snowflake is also extending trusted business definitions across the broader ecosystem, allowing external AI agents and major BI tools teams already use to access consistent, governed business context. Built on Snowflake’s continued commitment to openness, Horizon Context also supports the Open Semantic Interchange (OSI), making trusted business definitions universally accessible without vendor lock-in.

“In the financial industry, trusted data and consistent business context are critical to delivering accurate insights and managing risk across global markets,” said Jeff Miller, Managing Director, Global Head of Data Factory & Enterprise Data Platforms, BlackRock. “As AI becomes increasingly embedded across our enterprise, it’s essential that applications, analytics, and agents operate from the same trusted understanding of the business. Snowflake’s Horizon Context helps extend consistent business definitions across our broader data ecosystem, supporting more trusted and governed AI and analytics experiences at scale.”

Breaking the Security Barrier to Scalable AI

According to a recent McKinsey study, nearly two-thirds of organizations cite security as the top barrier to scaling AI⁵. This is because traditional access controls were built for human users, not AI agents capable of independently accessing systems, reasoning over sensitive data, and taking action across enterprise environments.

Snowflake is introducing new security capabilities that bring zero-trust security to the agentic era and help enterprises like Acxiom, NewDay, and Thomson Reuters strengthen security, visibility, and control as they scale AI. Built within Horizon Catalog, these innovations make Horizon the comprehensive trust layer for enterprise AI, connecting security, governance, observability, and business continuity into a cohesive platform for securing enterprise data, agents, and apps.

To help enterprises securely operationalize AI, these new security innovations deliver:

• A Security Model Built for Agent Access: Trusted AI requires a security model built for autonomous actors. Agent Identity¹ provides agents a verified identity before they can access enterprise data or take action, enforcing role-based permissions and maintaining a complete audit trail of every single agent activity. This enables enterprises with the visibility and control needed to securely deploy agentic systems at scale and prevent rogue agent actions.

• AI Security Posture Management for the Agentic Enterprise: As organizations deploy AI across the enterprise, maintaining visibility into security posture is critical. Innovations to Snowflake Trust Center now help organizations continuously monitor the security posture of AI systems, investigate violations faster, and accelerate risk response through AI-guided, context-aware assistance. This helps security teams stay ahead of emerging risks with greater visibility and control, while reducing alert fatigue as AI workloads scale across the enterprise.

• Enterprise Defense in the Agentic Era: As AI agents gain broader access to sensitive data, Snowflake provides centralized governance to prevent unauthorized exposure or manipulation. By enforcing consistent security policies across all AI workloads, Snowflake helps organizations neutralize threats like ransomware³ and data exfiltration³ while reducing the risk of compromised agents and costly business disruptions while scaling AI. Snowflake also delivers proactive, AI-native protection to stop emerging threats at machine speed. By integrating machine learning-driven detection with advanced prompt injection safeguards, enterprises can block jailbreak attempts and emerging zero-day vulnerabilities, helping secure enterprise AI without slowing innovation.

“As AI becomes increasingly embedded across the marketing industry, having the right security foundations in place is critical to our business scaling innovation responsibly,” said Ankur Jain, Chief Cloud and Data Modernization Officer, Acxiom. “Snowflake’s new AI security capabilities have the potential to provide greater visibility and control over how AI systems access and interact with personally identifiable data, helping us scale AI adoption responsibly while maintaining the trust our clients expect.”

“At Thomson Reuters, responsible AI adoption depends on strong security, visibility, and governance around how AI systems interact with enterprise data,” said Caitlin Halferty, Head of Data & Analytics, Thomson Reuters. “As AI becomes more deeply embedded in professional workflows and customer experiences, protecting sensitive information while enabling innovation is critical. Snowflake’s new AI security capabilities give us greater control and visibility, helping us scale AI with the trust, compliance, and accountability our customers expect.”

Make Enterprise AI Faster and More Efficient

Governance and security are often perceived as barriers to AI innovation that introduce complexity, slow access to data, and create operational friction just as organizations are trying to move faster. At the same time, AI introduces highly dynamic and unpredictable workloads that make managing compute at scale increasingly difficult. Combined with the connected governance, visibility, and control provided by Horizon Catalog, Adaptive Compute removes this complexity by automatically determining the optimal mix of compute and software resources in real time to deliver fast, efficient performance for AI and app workloads without manual tuning or infrastructure management. Together, Horizon Catalog and Adaptive Compute enable organizations to scale AI with a true serverless experience that combines consistent governance and security across data, AI, and compute with the speed, simplicity, and operational efficiency required to accelerate innovation.

Learn More:

• Learn more about how Horizon Context is creating a shared, trusted understanding of data for AI in this blog post.

• See how Snowflake’s AI security innovations are securing and governing AI at scale in this blog post.

• Double click into how Adaptive Compute is delivering performance, scale, and cost efficiency automatically in this blog post.

• Explore the latest innovations powering the foundation for trusted AI in this blog post.

• Check out all the innovations and announcements coming out of Snowflake Summit 26 on Snowflake’s Newsroom.

• Stay on top of the latest news and announcements from Snowflake on LinkedIn and X, and follow along at #SnowflakeSummit.
  
   

¹ Snowflake product is now generally available.

² Snowflake product is generally available soon.

³ Snowflake product is now in public preview.

⁴ Snowflake product is now in private preview.

⁵ McKinsey & Company, State of AI trust in 2026: Shifting to the agentic era. Based on McKinsey’s 2026

AI Trust Maturity Survey, which gathered responses from approximately 500 organizations globally between December 2025 and January 2026.