SAS Strengthens AI Trust with New Governance Tools
As enterprises move beyond experimentation to operational deployment of AI, concerns around control, transparency, and trust are growing. To address this challenge, SAS has introduced a suite of new products at its annual SAS Innovate conference designed to enable organizations to govern AI effectively while maximizing its potential.
The company announced offerings including copilots, an agent framework with Model Context Protocol (MCP) plugins, management tools, and more—all built on the Viya platform. According to Marie-Nellie Propre, global head of AI strategy at SAS: “We’re now seeing a critical shift from experimental AI to AI that actually acts and makes decisions autonomously.” This transition creates new requirements for trust, governance, and accountability.
Enhanced User Experience with Conversational AI
SAS Viya Copilot serves as an interactive assistant enabling users to manage analytical workflows through natural language. Integrated with Microsoft Foundry, it allows data analysis, model building, and decision-making tasks to be initiated via simple prompts—even by non-technical users.
“Copilots provide access to expert capabilities that empower users to explore the entire analytics lifecycle more easily,” explained Propre. The tool offers features like:
- Question answering across applications
- Generation of explainable AI code with documentation
- Guided model pipeline recommendations
- Interactive dashboards and visual analysis narratives
The company plans to extend copilot functionality to data management, model governance, and AI infrastructure in the future.
Connecting Agents and Ensuring Control
Beyond copilots, SAS introduced tools for managing both internal and external agents. The new SAS Viya MCP Server standardizes how models connect, enabling secure access to SAS resources from various LLMs like Claude, GPT, and Gemini without custom integrations.
“Copilots can now orchestrate complex tasks across multiple AI services,” Propre noted. Complementing this is the Agentic AI Accelerator—a collection of code samples, interfaces, and best practices that empower developers and citizen data scientists to build, deploy, and manage agents within SAS Viya.
Strengthening Governance with AI Navigator
SAS emphasized the need for robust oversight as AI expands. The new SaaS-based tool, AI Navigator, provides end-to-end visibility of all AI models—including both proprietary and third-party solutions—alongside governance controls and policy enforcement.
According to Reggie Townsend, VP of data governance at SAS: “AI Navigator not only shows what assets you have but also how well they’re governed.” He added that trust is becoming a key differentiator in the market, requiring organizations to demonstrate responsible AI practices.
Addressing Data Trust Issues
With AI increasingly reliant on high-quality data, SAS also announced enhancements to its Viya-based data management portfolio. These include features for:
- Data lineage tracking and transparency
- Built-in governance controls from design phase
- Agentic AI and copilot capabilities
- Cloud-native analytics acceleration
The company noted that this approach ensures data is both accessible and trustworthy, enabling organizations to confidently scale their AI initiatives.