A conference panel discusses AI, cloud computing, and data security before an audience in a San Francisco venue.
SAS spent Microsoft Ignite in San Francisco making a pitch that will sound familiar to anyone who has sat through an enterprise AI briefing: trust is the thing that separates AI projects that pay off from the ones that don't. This is a vendor-driven story, so it needs some unpacking. The facts are real and the numbers are interesting. The numbers also come from research that SAS sponsored.

A conference panel discusses AI, cloud computing, and data security before an audience in a San Francisco venue. What SAS actually did at Ignite​

At Ignite, held in San Francisco from Nov. 17 to 20, SAS said its experts would highlight how its partnership with Microsoft helps customers make governed decisions with their AI investments. That includes SAS Viya, the company's data and AI platform, on Microsoft Azure, plus its industry-specific solutions and decision intelligence. The company singled out financial services, life sciences and the public sector, where reliable results and clean data are operational requirements.

This was not a new product launch. It was a conference-presence and partnership story. SAS's own event page lists the concrete items:

  • Tuesday, Nov. 17, 4:30–5:20 p.m. PT (booth 5403): "Pure CPU Power: SAS Retrieval Agent Manager on Intel Xeon 6." SAS says the demo covers OCR, vectorization, embedding and retrieval. The pitch is that key stages of agentic AI pipelines can run efficiently on CPUs, which would reduce dependence on GPUs. That is SAS's description of its own demo. No independent benchmark was published alongside it.
  • Wednesday, Nov. 18, 11–11:30 a.m. PT (booth 5364L): "Advanced analytics without the complexity," a hands-on tour of SAS Viya across multiple data sources and algorithms.
  • Thursday, Nov. 19, 10:45–11:10 a.m. PT: "Apply Quantum AI to Complex Business Decisions." The session description says it covers how quantum AI may extend existing systems, and how SAS Viya on Azure and SAS Quantum Lab support use-case discovery and hybrid experimentation.

SAS also said booth 5364L would show Viya use cases in public health and in risk and fraud, including a synthetic-image detection demo aimed at fraudulent AI-generated images.

The quantum language is deliberately tentative. SAS's Amy Stout said the company is "exploring" where quantum can complement classical AI. That is not a claim of production-ready quantum capability.

The 15x statistic, and what it can and can't tell you​

The headline number comes from SAS's second annual Data and AI Impact Report: The New Economics of Trust. It was published with IDC research insights and sponsored by SAS. SAS's September 15 release says the findings rest on a survey of 2,699 decision-makers in 28 countries. They span four industries: banking, insurance, life sciences and the public sector.

The core finding is that organizations applying trustworthy-AI practices were 15 times more likely to report strong or high ROI. The release puts the figures at 62% versus 4%.

Some caveats are worth keeping in mind:

  • It's an association, not a proven cause. Well-run organizations may simply do both things well: govern AI and get value from it. The survey can't separate those effects.
  • ROI is self-reported. The respondents are people who know about or influence their company's AI initiatives.
  • Sponsorship matters. SAS sells governance, data and AI tooling. The framing that trust drives returns is also a sales message.
  • The scoring is SAS-defined. Organizations are scored out of 100 across five dimensions: data quality and governance, model governance and oversight, explainability and fairness, responsible-AI policy, and audit and accountability. Those averaging 80 or higher are classed as leaders.

IDC's Chris Marshall said trust is now a business imperative rather than just a compliance consideration. That is a statement of opinion attributed to him, not an independent test of the survey results.

Other findings IT teams may find useful​

Several other figures in the report speak to day-to-day AI operations:

  • Overrides are common. The release reports that 97.2% of respondents override AI-generated recommendations in at least some cases. The top reason was that the AI could not explain its decision, whether or not the output was correct.
  • Trust drops as autonomy rises. Reported trust was 76% for generative AI and 66% for agentic AI.
  • Data foundations lag. Only 17.5% of organizations reported fully optimized data infrastructure, up from 10.2% the year before. Organizations with optimized foundations were four times more likely to expect strong AI ROI.
  • Leaders keep spending. 85% of leaders were increasing trustworthy-AI investment by more than 10%, against 26% of laggards, according to SAS's report page.
  • Industry gaps are wide. In life sciences, 91% of leaders had responsible-AI policies reaching every employee, versus 7% of laggards. In banking, the figures were 85% and 29%.

For administrators, the practical reading is simple. If staff override AI because they can't see how it reached a conclusion, explainability and data lineage are adoption problems, not just compliance paperwork.

The Microsoft angle: Azure Marketplace and governance tooling​

SAS's Ignite page also promoted SAS AI Navigator, a governance product. It listed the product as available on Microsoft Marketplace with a six-month free trial. SAS describes AI Navigator as a SaaS solution that lets organizations inventory and govern AI use cases. It was first announced in April, with an Azure Marketplace target of Q3 2026.

Third-party coverage helps with context. One report says it launched on Microsoft Marketplace on October 5 and aims to give risk and compliance leaders a single inventory of the AI they use, with policies applied to each use case. The same coverage says it runs on its own and works for new or existing SAS customers. That means a Viya deployment isn't required.

One WindowsForum community write-up says the Marketplace listing requires Microsoft Entra ID and moves to pay-as-you-go pricing after the free period. It also notes the listing doesn't say the product is built into Microsoft 365, Azure, Copilot or Defender. That write-up is a community summary, not an official SAS or Microsoft statement. Check the listing's current terms before planning around it.

What this means for Microsoft-centric shops​

  • Procurement: Marketplace distribution can simplify buying. Whether a purchase counts toward an existing Azure commitment is not something the reviewed material establishes, so ask your Microsoft rep.
  • Overlap: Many organizations already use Microsoft's own compliance, identity and security tooling. Any third-party AI inventory has to fit alongside those controls, not duplicate them.
  • Evidence gap: SAS has not published customer results, a technical architecture for Viya on Azure, or independent validation of its CPU-versus-GPU claims. Treat these as claims to test in a pilot.
  • Survey use: The report's figures can help justify governance spending internally. They are best cited as "reported by SAS-sponsored research", not as proof that governance guarantees returns.

Bottom line​

SAS's Ignite message is coherent and consistent with a real industry problem: people don't use AI they can't understand, and many data estates aren't ready for autonomous agents. The 15x figure is striking but rests on a sponsored, self-reported survey. The actionable takeaways are modest: prioritize explainability, fix data foundations, and evaluate governance tooling on its merits. Nothing announced at Ignite changes the fact that the proof will be in your own deployment.


Correction (November 20, 2026): An earlier version of this article dated SAS's Data and AI Impact Report release September 1; the release is dated September 15, 2026.

 

References

  1. SAS Showcases Next Frontier of Trusted AI at Microsoft Ignite - HPCwire HPCwire Fri, 20 Nov 2026 18:59:51 GMT
  2. SAS showcases next frontier of trusted AI at Microsoft Ignite sas.com
  3. Organizations with trustworthy AI practices are 15 times more likely to see strong ROI, per study findings prnewswire.com