From Assistants to Agents: How SAS Is Bringing Agentic AI to Customer Intelligence

Jennifer_SAS — Wed, 25 Feb 2026 14:45:53 GMT

From Assistants to Agents: How SAS Is Bringing Agentic AI to Customer Intelligence

AI has moved quickly from novelty to necessity in marketing. But as capabilities accelerate, so do the questions: How autonomous should AI be? Where does human control matter most? And how do we move from isolated AI features to systems that genuinely improve marketing outcomes?

At SAS, our answer is clear: AI must work alongside marketers, not replace them. In SAS Customer Intelligence 360 (CI360), we are building a new generation of specialised AI agents that collaborate with users, operate within clear guardrails, and deliver real value across audiences, journeys, decisioning, and execution.

Why Agentic AI Matters Now

Agentic AI represents a step change from traditional assistants and copilots. Rather than responding to isolated prompts, agents can plan, retrieve context, take actions, and coordinate with other agents to achieve defined goals.

History shows us that progress in AI often feels gradual - until it isn’t. As seen in domains like chess, systems improve steadily for years and then suddenly surpass human performance in narrow tasks. The lesson for marketing is simple: we need to take agents seriously but deploy them responsibly.

That’s why SAS is focused on agentic capabilities that:

Operate inside the frontier of trustworthy AI

Maintain humanintheloop (HITL) control

Are deeply embedded into real marketing workflows, not bolted on later

A MultiAgent Future for CI360

Rather than a single, monolithic AI, SAS is building a multiagent system within CI360. Each agent is specialised, contextaware, and designed to operate on behalf of the marketer while remaining transparent and controllable.

In this model, agents don’t replace existing capabilities such as audiences, journeys, destinations or marketing decisioning. Instead, they operate across them, retrieving the right objects, making recommendations, and accelerating execution under human supervision.

This approach allows CI360 to evolve from a powerful platform into an intelligent operating system for customer engagement.

The Journeys Agent: From Idea to Execution

One of the most significant steps forward is the Journeys Agent, coming in Q2.

The Journeys Agent allows marketers to create journeys using multimodal inputs including text briefs, images and conversational prompts. The agent automatically retrieves relevant audiences, events and touchpoints, then assembles a journey structure that reflects both best practice and the marketer’s intent.

Crucially, this is not “blackbox” automation. Humanintheloop checkpoints are built into both the planning and creation stages, ensuring marketers stay in control while dramatically reducing build time and complexity.

Behind the scenes, the agent generates accurate and consistent SAS code, ensuring journeys are not only faster to create but also productionready from day one.

The SAS 360 Agent: Coordinating Intelligence Across CI360

As more agents come online, coordination becomes critical. This is where the SAS 360 Agent plays a central role.

The SAS 360 Agent acts as a supervisory layer, managing interactions between specialised agents such as Audience, Journey, Email, Search and Recipes agents. Rather than forcing marketers to jump between tools, the system works alongside users, orchestrating actions across customer data, marketing AI and journey execution.

The result is a system that feels less like a collection of features and more like a collaborative marketing partner.

Explainability Built In, Not Bolted On

As AI takes on more responsibility, explainability becomes nonnegotiable.

Within CI360, explainability is embedded directly into Marketing AI. Marketers can understand why a model behaves the way it does, how decisions are made, and what data influenced the outcome without needing to become data scientists themselves.

This focus on transparency ensures AIdriven decisions remain accountable, auditable and aligned with both regulatory expectations and brand values.

The Search Agent: Answers When You Need Them

Another practical example of agentic value is the Search Agent, also planned for Q2.

The Search Agent allows users to ask operational and performancerelated questions across their CI360 tenant starting with tasks and audiences and progressively expanding to additional object types. Instead of navigating menus and dashboards, marketers can simply ask and receive contextual answers grounded in their own data.

It’s a small shift in interaction, but a big leap in usability.

Agentic AI, Delivered the SAS Way

What makes SAS’s approach to agentic AI different is not ambition alone but discipline.

By building agents that are:

Embedded across the full CI360 experience

Designed with human oversight as a first principle

Governed, explainable and enterpriseready

SAS is helping marketers move faster without sacrificing trust.

Agentic AI isn’t about handing control to machines. It’s about creating systems that amplify human expertise one specialised agent at a time.

Re: From Assistants to Agents: How SAS Is Bringing Agentic AI to Customer Intelligence

ClaudiaSpermann — Mon, 02 Mar 2026 10:46:52 GMT

Looking really forward to the Journeys Agent.

topic From Assistants to Agents: How SAS Is Bringing Agentic AI to Customer Intelligence in SAS Customer Intelligence

From Assistants to Agents: How SAS Is Bringing Agentic AI to Customer Intelligence

Re: From Assistants to Agents: How SAS Is Bringing Agentic AI to Customer Intelligence