SAS Unveils AI Governance Tools to Control Agentic AI in Enterprise

At SAS Innovate in Grapevine, Texas, SAS rolled out a suite of AI governance tools designed to help enterprises manage the growing complexity of agentic AI systems. The announcements center on SAS AI Navigator, a new governance product, alongside expanded capabilities in the SAS Viya platform and SAS Customer Intelligence 360.

The timing reflects a broader industry tension. Enterprise AI agents are moving faster than many organizations can govern them, creating what Reggie Townsend, VP of SAS AI Ethics, Governance and Social Impact, calls a trust dilemma. Nearly half of AI users like the promise of automation but hesitate to turn over control to autonomous systems due to risk exposure.

"AI governance is too often thought of as a compliance measure, when it is actually a growth driver," Townsend said during the conference. The company wants to position governance not as a brake on innovation but as the foundation that lets businesses use AI agents without handing them the keys unsupervised.

SAS AI Navigator arrives as a Software-as-a-Service solution available in Q3 2026 on Microsoft Azure Marketplace. The tool compiles an AI inventory and aligns AI use cases with government regulations and internal policies. Companies using chatbots to interact with customers can govern agents or models such as Claude or Microsoft Copilot and apply policies to ensure regulatory compliance.

According to the official SAS Data Management press release, the platform embeds governance, lineage, and performance directly into data workflows. This approach treats auditability as a core design principle rather than a compliance afterthought layered on top of disconnected tools.

The physical reality of this matters. Instead of navigating multiple dashboards to verify data lineage, users can trace how data moves through the system without leaving their workflow. It's the difference between checking a spreadsheet for errors versus having the spreadsheet flag issues as you type (a small friction reduction that compounds across teams).

SAS Viya has been expanded with governed AI assistants and agentic AI capabilities. These tools help business and analytics teams move from experimentation to production-ready intelligence. Jared Peterson, senior vice president of Global Engineering at SAS, emphasized that human expertise gets elevated by automation, not diminished by it.

With SAS Viya, organizations can pair copilots and agents with human judgment, trusted data, and enterprise governance. The goal is ensuring AI doesn't just generate outputs but drives responsible, real-world decisions. This distinction matters when the alternative is shadow AI operating outside organizational oversight.

Industry-specific agents are also part of the rollout. The SAS Supply Chain Agent streamlines supply and operations planning for retailers and manufacturers. Someone can use this agent to run a scenario such as a sudden 15% drop in demand. The agent explores possible outcomes, explains what is driving the change, and supports recommendations by showing how it arrived at its decisions.

SAS Customer Intelligence 360 has added agentic AI capabilities with specialized agents working alongside marketers. Rather than a single, monolithic AI, a multi-agent system has been evolved within the platform so that specialized, context-aware agents handle tasks tailored by the marketer.

Human-in-the-loop control is embedded directly into marketing workflows. The SAS 360 Agent serves as a supervisory layer, managing interactions between specialized agents such as Audience, Journey, Email, Search and Recipes agents. Marketers don't navigate multiple tools and interfaces; the agent orchestrates actions across customer data, marketing AI and journey execution.

The Journeys Agent helps marketers create customer journeys using multi-modal inputs, including text briefs, images and conversational prompts. Human-in-the-loop checkpoints are embedded throughout the planning and creation process. Behind the scenes, the agent generates consistent, production-ready SAS code.

According to SAS's official announcement on Customer Intelligence 360, this approach transforms the platform from a powerful marketing tool into an intelligent operating system for customer engagement. Marketers no longer must choose between relevance and efficiency.

The governance push responds to documented barriers. Research from IDC and SAS found nearly half (49%) of organizations cited noncentralized or poorly optimized cloud data environments as the top barrier to AI progress. Insufficient data governance processes followed closely at 44%. Gartner predicts that 60% of AI initiatives will fail due to a lack of AI-ready data.

SAS Data Management has been refreshed to make it ready for agents and automation to operate within it. By embedding governance, lineage, and performance directly into data workflows, confidence in AI outcomes increases without sacrificing trust or control. The platform brings analytics and AI directly to the data, wherever it resides, rather than moving data across platforms.

This reduces unnecessary data movement, improves performance and preserves the lineage and auditability organizations need to trust their results. SAS SpeedyStore, a high-performance, cloud-native analytical data platform tightly integrated with SAS Viya, runs analytics and AI alongside distributed data.

Looking further ahead, SAS Quantum Lab is coming in Q4 2026. The purpose is to reduce the cost of quantum AI exploration and help users avoid false signals as they explore this technology credibly. One of the most promising immediate use cases is enhancing the accuracy of fraud detection systems in financial services by enabling more efficient identification of complex transaction patterns.

The announcements come as enterprises face growing concerns about shadow AI, compliance exposure, and the trustworthiness of autonomous systems in production. TechRepublic covered the event, noting that SAS wants to position governance as the foundation that lets businesses use AI agents without handing them the keys unsupervised.

Whether organizations actually adopt these tools at scale remains the real question. Governance infrastructure is only useful if it's convenient rather than a drag on business and innovation velocity. Townsend acknowledged this tension explicitly during the conference.

"AI is great for optimization but lacks judgment when it comes to dealing with ambiguity and competing values," he said. The tools SAS is launching attempt to bridge that gap, but the market will decide if the trade-off between control and speed is worth it.

Artūras Malašauskas is an AI Systems Integrator with 20+ years of production-grade web engineering experience. He has designed, shipped, and scaled enterprise Python/PHP systems for logistics, SaaS, and public-sector clients. For the past year, he has focused exclusively on AI integrations: deploying open-source LLMs, building generative media pipelines (image, audio, video), and engineering multi-agent workflows for real production environments. His standard: reproducibility, security, cost-efficient inference—no vaporware. He documents and evaluates emerging AI tooling, separating verified capabilities from marketing noise. Technical editor at: muza-ai.eu, ai-verslas.lt, ai-naujinos.lt Connect on LinkedIn