(Bryan Harris, EVP, CTO, SAS, on stage at SAS Innovate 2026. Source: SAS.)
SAS (Cary, N.C.) EVP and CTO Bryan Harris opened his keynote at the company’s SAS Innovate 2026 conference in Grapevine, Texas with a stark question: “Three words. Will people matter?”
The framing set the tone for a speech that positioned the current wave of artificial intelligence not as a technological crisis, but as a test of how organizations define the role of human judgment in an increasingly automated world.
“We are in a crisis. A crisis of confidence, of human ingenuity,” Harris said. “It’s not a collapse in the belief that AI will matter. It’s a collapse in the belief that people will matter.”
For Harris, the stakes were immediate and strategic. “The way you show up right now with AI will determine your brand value to the next generation,” he said, adding that “it’s not about choosing between people and profit.”
Photo by author. (Click to enlarge.)
Instead, he framed AI as the latest in a series of technologies that extend human capability. “AI will scale human observation and decision making,” he said, placing it alongside earlier advances that scaled knowledge, connectivity, and computation.
Closing the Information Gap
Harris grounded the argument in the widening gap between data creation and human capacity to interpret it.
Organizations, he said, are creating “exponentially more data every day with no end in sight,” producing an “information landscape” that can overwhelm the workforce. The result is “information overload that fuels confusion, fatigue, and opportunity.”
The role of AI, in this formulation, is to help close that gap by scaling human observation and decision-making—an extension of SAS’ long-standing analytics mission.
Agentic AI and the Problem of Trust
A central focus of Harris’ remarks was the emergence of agentic AI, which he framed in terms of variability and control rather than novelty alone.
He distinguished between deterministic systems—where “given the same input, it produces the same output. Every time”—and non-deterministic systems based on large language models, which “can produce different outcomes from the same input.”
In a financial crimes example, he noted that “three different runs can produce three different outcomes,” depending on context and interpretation.
Photo by author. (Click to enlarge.)
At the individual level, this variability can be managed through human oversight. At the enterprise level, however, trust must be designed into the system.
“Trust shifts from the individual to the accuracy and repeatability of the process itself,” Harris said.
The implication is that agentic AI requires embedded controls. “You must add guardrails to ensure accuracy and prevent compound error at scale,” he said.
From Coding to ‘Context Engineering’
To operationalize this, Harris outlined a four-phase model for AI-enabled workflows: design, execute, verify, and validate.
In the design phase, he said, “human value shifts from typically writing code to the new discipline of context engineering,” requiring domain expertise to define requirements and objectives.
Verification ensures systems meet technical specifications, while validation ensures alignment with business objectives. “The verification and validation phases are absolutely critical for guardrails to ensure accuracy,” Harris said.
High-Stakes Use Case: Clinical Trials
Harris illustrated the model with a life sciences example, describing how AI can be applied to clinical trial design and execution.
Photo by author. (Click to enlarge.)
AI agents can read and interpret trial protocols, map them to regulatory standards, and process large volumes of clinical data. Verification and validation processes can reduce inconsistencies and accelerate timelines.
He said this approach can result in “a clinical trial submission that runs in weeks rather than months.”
But he emphasized the risks of failure: misinterpreted protocols or inconsistent submissions can halt trials, delay therapies, and increase costs—underscoring the importance of governance in high-stakes environments.
Convergence and the Rise of Digital Twins
Harris placed agentic AI within a broader convergence of technologies, including machine learning, generative AI, computer vision, synthetic data, and optimization.
“When these capabilities come together, you don’t just get better models, you get a digital twin,” he said, describing a dynamic representation of an organization’s operations.
He pointed to work with a Danish medical sterilization provider, where a digital twin helped identify workflow bottlenecks and train computer vision models using synthetic data.
“All of these improvements ultimately mean that doctors get the tools they need, surgeries start on time, and patients get the care they expect,” Harris said.
Architecture as Continuity
Bryan Harris, EVP, CTO, SAS on stage at Innovate 2026. Source: SAS. (Click to enlarge.)
In a recorded conversation presented during the keynote, SAS founder and CEO Jim Goodnight reflected on early architectural decisions that continue to shape the company’s technology.
Goodnight described the development of a multi-vendor architecture in the 1980s, designed so that software could run across different computing environments without being rewritten for each one.
Harris connected that approach to current multi-cloud and emerging architectures, suggesting continuity in SAS’ design philosophy over time.
A Culture Built Around People
The Goodnight segment also reinforced a cultural theme that Harris connected to the AI era: the primacy of people.
Harris referenced a long-standing company philosophy that places users at the top of the organizational structure and leadership in a supporting role, emphasizing the importance of enabling employees to serve customers effectively.
“My people are my greatest asset,” Goodnight said, describing a management approach that has shaped SAS over five decades.
Beyond the Moment
Harris closed by returning to his opening question. “Will people matter? Of course we will,” he said. “In fact, we have never been more important than we are right now.”
He described AI as part of a broader pattern in which transformative technologies reshape society before becoming part of everyday infrastructure.
“The only thing that outlasts our innovation is people,” Harris said.