SAS pushes deeper into agentic AI

Global data analytics pioneer SAS has expanded its Viya platform with new agentic AI features. The advancements include AI assistants, agent infrastructure and acceleration tools.

The updates build on Viya’s existing AI foundation, which the company says can help organisations move from isolated generative AI (GenAI) use cases to governed, production-ready intelligence at scale.

The news was revealed during SAS Innovate 2026, the company’s global data and AI conference, in Dallas last month (April).

SAS provides the following new features:

• SAS Viya Copilot – A family of AI assistants embedded across the analytics life cycle.
• SAS Viya Model Context Protocol (MCP) Server – Using the open MCP standard to expose SAS Viya analytics and decisioning capabilities as tools for AI agents.
• SAS Agentic AI Accelerator – A curated framework for building, governing and deploying AI agents within SAS Viya.

“The role of human expertise in operationalising agentic AI is not diminished by automation; it’s elevated,” says Jared Peterson, SAS SVP of global engineering. “With SAS Viya, organisations can pair copilots and agents with human judgment, trusted data and enterprise governance, so AI doesn’t just generate outputs but drives responsible, real-world decisions.”

Copilot for data and AI tasks

SAS Viya Copilot is a governed, conversational AI assistant embedded directly into the platform. The system is designed to work alongside human experts inside production-grade analytics workflows. SAS says Viya Copilot can help organisations analyse data, build models and make decisions across the analytic life cycle while maintaining enterprise-grade security and oversight.

SAS Viya Copilot integrates Microsoft Foundry to operate inside analytics workflows rather than as a standalone chat interface. According to the company, Copilot helps data scientists, developers and business analysts generate, explain and act on insights using natural language across both coding environments and interactive visual interfaces.

SAS provides the following capabilities for Viya Copilot:

• General questions and answers across core Viya applications. This includes data discovery, model pipeline development, model management, decision intelligence, and environment management.
• Code acceleration with AI-generated SAS and Python code, documentation and explanations.
• Model pipeline guidance with intelligent recommendations, next steps and explainability.
• Conversational dashboarding with AI-driven data enrichment, natural language dashboard creation and automated insights.
• Visual investigation with AI-assisted search and AI-powered case and alert narratives.

Over time, says SAS, Viya Copilot data and AI capabilities will extend across data management, model management and AI infrastructure, enabling consistent AI assistance from data ingestion to deployment.

According to SAS, Viya Copilot is designed to execute domain-aware analytics tasks grounded in industry data models, terminology and best practices, unlike many horizontal AI assistants that rely primarily on pretrained LLM knowledge or generic tools. 

SAS provides the following industry Copilot capabilities:

• SAS Asset and Liability Management (ALM) Copilot, which guides analysts through scenario configuration, execution and interpretation for financial risk workflows, translating natural-language inputs into governed analytic models.
• SAS Health Clinical Data Discovery Copilot, which accelerates clinical data discovery and analysis by guiding users through data exploration, cohort creation and quality investigation using natural language.

The company plans to expand Viya Copilot for additional industries in 2026, including financial crime prevention in banking, and intelligent planning and supply chain optimisation for manufacturing.

From copilots to agents

While SAS Viya Copilot can improve how people interact with analytics, agents allow those capabilities to compound by connecting insights, decisions and actions across systems at scale.

Beyond embedded AI assistants, SAS Viya can provide the infrastructure and tools organisations need to build, connect and govern AI agents at scale. These capabilities, available to current Viya users through GitHub, are designed to help transform isolated GenAI use cases into enterprise-ready agentic AI that works across systems, teams and business processes.

SAS Viya Model Context Protocol (MCP) Server can provide a standardised way to expose SAS analytics, models and decisioning capabilities to external AI agents. By using SAS Viya MCP, organisations can embed SAS’ analytics, governance and domain intelligence into their own agents from the LLM interface of their choice, such as Claude, without duplicating logic or bypassing enterprise controls.

SAS Agentic AI Accelerator can enable teams across different skill levels (no-code, low-code and developers) to design and operationalise AI agents with governance and trust. The accelerator provides a curated collection of code, components, interfaces and best practices to help teams build, govern, and deploy AI agents within SAS Viya, moving from experimentation to production in a repeatable, governed way.

SAS Retrieval Agent Manager (RAM) is a no-code solution built on the retrieval augmented generation (RAG) framework that aims to streamline the process of transforming raw unstructured data into context-aware AI responses. While currently only available as a standalone product, SAS plans to integrate RAM capabilities into SAS Viya, further enabling agents and assistants to ground responses in an enterprise context for accurate, trusted decisioning.

Together, says SAS, the updates position Viya as a governed, enterprise-ready platform for agentic AI, supporting AI assistants and intelligent agents that work together to compound value and turn trusted insights into real-world decisions at scale.

* To learn more about SAS’s Agentic AI products, visit the website here.

* Read Arthur Goldstuck’s coverage of SAS Innovate 2026 here.