AI Governance: From Compliance to Growth Lever - The Strategy of SAS and AI Navigator
AI Governance: From Compliance to Growth Lever - The Strategy of SAS and AI Navigator
In 2026, AI in businesses ceased to be an isolated project. In more mature organizations, proprietary models coexist with machine learning services from hyperscalers, agents built on commercial LLMs, and prototypes born in individual departments. The problem is no longer starting initiatives but keeping them together: knowing how many there are, where they operate, with what data, and under which rules. The spread of shadow AI, or the adoption of tools outside the IT perimeter, has become a real concern for those leading the technology function.
Gartner estimates that by 2030, over 40% of companies will experience security or compliance incidents linked to the unauthorized use of these technologies. A study conducted by SAS in collaboration with IDC adds another element: the adoption of agents and language models is outpacing the investments necessary to make them reliable.
In this context, AI governance risks being perceived as a brake, a compliance measure that slows down innovation. This is the viewpoint SAS seeks to overturn.
"AI governance is often seen as a compliance measure, but it is actually a growth engine," summarizes Reggie Townsend, Vice President of SAS AI Ethics, Governance and Social Impact.
The reasoning is that a clear framework of rules, transparency, and security does not limit the use of technology; it enables it: allowing organizations to move beyond pilot projects without exposing themselves to legal, reputational, or operational risks. It’s the difference between a company that knows where its AI stands and one that suffers its consequences.
Behind this statement lies a precise definition of what AI governance means for SAS: the set of practices by which an organization accelerates innovation, manages risk, and ensures that systems are trustworthy. Not a single control but a strategy that begins with internal culture and translates into operational tools for transparency, compliance, and systematic oversight. Companies that build it robustly defend the trust of customers, authorities, and boards, while those that forego it expose themselves to penalties, image damage, and opaque decisions that are difficult to account for.
A Competency Built in Fifty Years
SAS approaches this topic with a long history. Founded in 1976 and still controlled by its founders, the company has made model traceability and risk management one of its hallmarks, long before generative AI brought the issue to widespread attention. The SAS Viya platform has long integrated model risk management functions, result interpretability, bias detection and mitigation, data protection, and continuous monitoring throughout the entire lifecycle.
On this front, the company has achieved significant external recognition: Chartis Research has placed it among the category leaders in its RiskTech Quadrant dedicated to AI governance solutions.
"By combining these capabilities with deep experience in regulated sectors, SAS is positioned to present AI as a growth strategy for its clients," noted Michael Versace, Research Director at Chartis.
AI Navigator does not emerge in isolation but as a piece of a portfolio that already includes tools such as SAS Model Manager and SAS Model Risk Manager.
A Unique Census, Even for AI from Others
SAS AI Navigator is a SaaS service set to launch in the third quarter of 2026 on the Microsoft Azure Marketplace. It is designed for those involved in AI, data, compliance, and risk, allowing for the creation of a unique census of all AI use cases present in the company, the models and agents that power them, and the rules applied to them. What distinguishes it from competing proposals is that governance does not stop at the models developed by SAS. A corporate chatbot based on Claude or Microsoft Copilot, to name two examples mentioned by the company, falls within the same policy perimeter applicable to internal models of the platform. There is no need to change how AI is built: AI Navigator offers a unified view of predictive models, generative AI, and agency systems, whether developed internally or purchased from third parties.
The initial choice is the most interesting. Many governance tools focus on technical artifacts, i.e., individual models. AI Navigator, however, starts from the use case, the point at which AI generates an effect on the business, linking ownership, purpose, policies, and responsibilities. From a single interface, designed to be readable even by top executives without programming expertise, all use cases can be viewed, alerts on documentation gaps and priority risks received, initiatives can be aligned with external regulations through importable rule packages, and a verifiable trail maintained for audits and reviews. The service integrates with SAS Viya but can operate independently alongside tools from other vendors and is delivered as SaaS to reduce the implementation burden on IT: elements designed to lower the entry barrier and not turn governance into a standalone project.
The message, articulated by Townsend, is that the tool must first and foremost be used:
"The greatest risk of an AI governance program is not regulation but adopting such a complex tool that no one ends up using it."
This position speaks volumes about SAS's strategy. While hyperscalers control the infrastructure level and large application vendors integrate copilots within their systems, there remains an uncovered horizontal layer—the cross-control of models and agents already in use, regardless of who produced them. This is the space SAS seeks to occupy, consistent with an approach the company has long claimed: starting from the problem and its implications, not from the technology. For regulated sectors, from banking to healthcare to public administration, where every automated decision must be explainable and documented, a unique and verifiable registry of the AI in use is not an accessory. It is the condition for truly using it.