SAS Viya launches agentic AI framework for ethical decisions
SAS has introduced a new framework for customisable AI agents on its SAS Viya platform, focusing on transparent, governed and ethically calibrated decision-making for enterprise customers.
The announcement details the SAS Viya agentic AI framework, which enables organisations to design, deploy and manage AI agents with oversight on both human and AI autonomy. According to the company, this system aims to create collaborative environments where AI and human decision-makers work together based on the complexity and risk of tasks.
Nick Patience, Vice President and Practice Lead, Artificial Intelligence Software and Tools at The Futurum Group, said: "SAS' approach to agentic AI strikes the critical balance between autonomous decision-making and ethical governance. Its intelligent agents represent not just technological advancement but a pragmatic framework for responsible enterprise AI adoption – precisely what organizations need as they navigate this rapidly evolving landscape to gain a competitive advantage."
The structure of the SAS Viya agentic AI framework rests on three main pillars. The first is a hybrid decisioning method, combining deterministic analytics with large language models (LLMs) to offer reliable and precise decision outputs. This approach, the company says, allows AI agents to function effectively under the necessary business rules and regulatory requirements that are essential in highly regulated sectors.
The second pillar focuses on balancing the degree of autonomy between human input and AI. The platform lets organisations determine the level of control AI agents have for different kinds of tasks, distinguishing between those that are suited to fully autonomous operation and those where ongoing human oversight remains essential for ethical judgement and strategic guidance.
The third aspect is governance, with Viya's embedded framework intended to ensure that output not only meets performance standards but also supports data privacy, organisational ethics and wider regulatory compliance.
Tiffany McCormick, Research Director, Digital Business Models and Monetization at IDC, stated: "As organizations evolve toward open, interoperable AI ecosystems across multi-cloud and hybrid environments, trust and explainability in AI governance are emerging as key differentiators among tech vendors. SAS is taking industry-leading steps to address the growing demand for agentic AI, with a clear commitment to ethical rigor and differentiated execution in AI decisioning."
SAS Viya is presented as supporting organisations throughout the entire lifecycle of building and maintaining AI agents—from initial data ingestion and processing, through to deployment, monitoring and adaptation. The framework provides mechanisms for continuous oversight, including audit trails, bias detection and compliance checks.
The company outlined future plans for its agentic AI roadmap, which include introducing co-pilot productivity assistants into the Viya ecosystem, aimed at helping users expedite their workflows while remaining aligned with enterprise requirements. There are also plans to release pre-packaged, domain-tailored AI agents for use in specific scenarios such as data engineering and supply chain optimisation. These domain-specific agents are designed to ease adoption and reduce time-to-value for customers without eroding their confidence in the automation.
Marinela Profi, Global AI Market Strategy Lead at SAS, commented: "SAS Viya builds agents that don't just act – they decide with purpose, guided by analytics, business rules and adaptability and grounded by decades of SAS' trusted governance. SAS' unified, governed, decision-first framework turns AI agents from a science experiment to a business differentiator."
SAS said it aims to position its platform as a comprehensive and regulated solution for organisations looking to integrate agentic AI into their digital workflows while maintaining confidence and clarity in AI-powered business decisions.
