SAS Innovate on Tour Sydney report: where human ingenuity meets AI

Enterprise AI is moving from experimentation to operational reality. The question is no longer whether to adopt AI, but how to deploy it in ways that create measurable business value while maintaining transparency, accountability and human oversight.

At SAS Innovate on Tour Sydney, business and technology leaders gathered to explore what this next chapter looks like. They heard how healthcare, financial services, manufacturing and logistics organisations are applying AI to solve increasingly complex challenges.

Yet as AI capabilities accelerate, leaders are recognising that technology alone is not enough. Unlocking long-term value requires the right balance of human expertise, intelligent systems and governance that inspires confidence across the enterprise.

The event highlighted the three themes shaping enterprise AI: the enduring importance of human ingenuity, the emergence of digital twins as the next evolution of intelligent systems and trust as the foundation for responsible AI adoption.

Together, these themes illustrate how organisations are moving beyond isolated AI applications to build intelligent, governed systems that empower people to make better decisions with greater confidence.

The importance of human ingenuity

Human ingenuity has never been more important than it is today. Human-created innovations — from the printing press to the internet — have driven every technological revolution.

“From the very beginning, humans have been the engine behind every breakthrough that has changed the world — not AI, but human ingenuity,” said Bryan Harris, Executive VP and CTO, SAS.

We are undoubtedly at a pivotal moment as AI reshapes workplace systems, but humans are at the centre of this next phase of innovation. AI will scale human observation and decision-making, Harris explained.

Whether it’s finding cures for cancer, stabilising the climate, expanding access to healthcare and housing, or stopping fraud, there are some big problems facing humanity. AI is poised to empower humans to tackle these challenges. The power comes through harnessing technologies to work together.

“Agentic AI is just one capability in the entire toolset alongside machine learning, generative AI, optimisation, computer vision and synthetic data. The value comes when these are orchestrated together,” said Harris.

In its 50-year history, SAS has pursued technology innovation to give people powerful insights into data and help them make better decisions.

“It’s always been about empowering people with technology to scale human observation and decision-making. Since the beginning, SAS has been pioneering technology breakthroughs to help you close this information gap, make better decisions and create a competitive advantage,” Harris told the attendees.

Today, organisations are generating exponentially more data, and with more intelligent, internet-connected systems, this will only continue. “This is the information landscape over time and volume of data, but the problem is that this information is overwhelming for the workforce to consume and make sense of,” he said.

The growth in data is being outpaced by the growth in our ability to process and make sense of that data. “The difference in these growth rates is the information overload that fuels confusion and fatigue, but this is also opportunity. So the question becomes: How do we close the gap?”

AI has been integrated into the SAS Viya platform to support enhanced decision-making for many types of organisations from finance and medical to research and logistics.

The philosophy shaped the development of Medibank’s Recover Boost product. Rather than beginning with generative AI, the team first asked what business decision needed to be made. The result is a real-time eligibility engine built on deterministic decision logic that delivers consistent outcomes while supporting members during some of life’s most challenging moments.

The technology supports staff in delivering faster, more consistent outcomes for members, while ensuring critical decisions remain transparent and repeatable.

Jonathan Butow, Head of Innovation at SAS, said organisations often ask the wrong question when implementing AI. “The question that people always ask is, ‘Where do we put the model?’... actually the sharper question is, ‘What’s the decision that we need to make?’”

In Medibank’s case, the focus is designing technology around the needs of the members. “We’re supporting our members in a really unique manner that we’ve never been able to do before at Medibank,” said Alicia Faour, Product Innovation Lead, Medibank.

The conference highlighted how organisations are moving beyond viewing AI as a replacement for people. Instead, the technology is becoming a tool that amplifies human expertise, helping employees navigate growing complexity while leaving the decisions that matter most in human hands.

SAS understands that AI will scale human observation and decision-making, and this won’t replace the need for people.

“Every breakthrough technology fades into the background; the only thing that has outlasted them all is people,” said Harris.

The next evolution of AI is digital twins

Digital twins are virtual replicas of businesses that combine multiple technologies to enable organisations to ask and answer increasingly sophisticated questions about their operations.

Machine learning helps predict outcomes, generative AI is used to interact, reason and explain, and agentic AI allows systems to plan and act across complex workflows. Digital twin technology also includes computer vision that creates new signals and detects patterns, synthetic data to safely explore scenarios that are risky or haven’t happened yet, and optimisation to choose the best outcome from all of the technologies.

Bryan Harris described digital twins as the next stage of AI evolution.

“When these capabilities come together, you don’t just get better models; you get a digital twin, a living replica of your business that allows you to ask and answer your most sophisticated questions,” Harris explained.

The attendees heard about how Georgia‑Pacific combined thousands of sensors, physics-based models and AI to create digital twins of manufacturing operations. Olivier Debaillon, Head of AI for ANZ at AWS, explained that Georgia‑Pacific used the photorealistic digital twin of its manufacturing facility to optimise vehicle routes and fleet sizes.

Engineers no longer have to manually assess every possible variable. AI explores thousands of potential scenarios, allowing experts to focus on selecting the best course of action.

“Instead of the subject matter expert looking at all the different scenarios, AI can explore thousands of scenarios and bring back the best ones for the expert to review,” explained Debaillon.

Digital twin technology has applications far beyond factory floors, from helping banks understand how geopolitical instability might affect balance sheets to modelling supply chain disruptions caused by extreme weather, shipping constraints or global conflict. Rather than reacting to events, organisations will increasingly be able to test multiple futures before deciding how to respond.

Digital twins are not replacing human expertise but extending it. As Harris observed, simulation enables leaders to ask increasingly sophisticated questions about uncertainty, resilience and risk, giving decision-makers a richer understanding of the consequences of different choices before they act.

The result is a shift from predictive AI to prescriptive decision-making. As simulation technologies mature, digital twins are poised to become an essential tool for organisations navigating increasingly complex operating environments, helping leaders move beyond forecasting what might happen to understanding how different decisions could shape the future.

Rather than relying solely on historical data, digital twins allow organisations to simulate future events, evaluate potential outcomes and make more informed strategic decisions.

As AI continues to evolve, digital twins promise to become an essential decision-support tool, helping leaders prepare for disruption before it occurs rather than simply reacting to it.

It will enable business leaders to ask questions that haven’t been possible to ask and answer them about the future. “How do I take complex systems with lots of dependencies and ask questions of it and have a series of outcomes that say: based off these conditions we think the best decision for you is X,” said Harris. “At SAS, we’re building the foundation for that future to exist.”

Trust as the currency of AI in business

The one ingredient that makes AI truly enterprise‑ready is trust. It is the foundation that supports confidence in AI-driven decision-making and encompasses explainability, accountability and human oversight.

“Leaders want the best of what AI has to offer without losing control over what matters most, yet our report found 46% of companies worldwide today are experiencing what’s known as a trust dilemma and as a consequence, they’re leaving 50% of their AI potential completely untapped,” said Reggie Townsend, Vice President Data Ethics Practice, SAS.

From model development and bias monitoring to governance and oversight, trust needs to be embedded across the AI lifecycle to ensure the best outcomes. Trustworthy AI isn’t just a requirement — it’s becoming a competitive advantage.

However, many organisations are not yet at the point of fully trusting AI and the benefits remain unrealised. As AI moves from experimentation to full enterprise deployment, business leaders will need to have faith in these systems.

“Organisations need the efficiency and productivity of AI without losing clarity and consistency, and we can’t overlook the financial: we need to capture value without creating unmanageable risk,” Townsend said.

Participants at the SAS Innovate on Tour conference in Sydney heard how organisations such as AustralianSuper, Medibank and UOB are adopting AI while managing risk. Barnali Das, Senior Manager Model Risk, AustralianSuper, highlighted the importance of robust model risk management and governance, emphasising the need for ongoing validation and oversight as AI becomes embedded in decision-making.

“We’re building a host of AI agents that can do the methodology research and implementation checks — compare different approaches and identify gaps — so our humans can focus on the judgmental side and take accountability for the work we do,” Das explained to the attendees. “Traditional governance has given us a foundation, but the AI evolution is moving us from periodic, control‑based oversight to something that is continuously monitored and genuinely risk‑based.”

Anselmo Marmonti, SAS Vice President of Risk, Fraud and Compliance Solutions, explained how organisations need governance frameworks that evolve alongside AI, enabling innovation while maintaining appropriate controls.

“Our perspective is really to focus on the business processes needed to achieve the goals and understand how to re‑imagine those processes — how AI agents and generative AI can help achieve these goals while reducing cost,” he said.

Responsible AI requires continuous oversight across the AI lifecycle, rather than a one-off validation exercise before deployment. Marmonti also reinforced the idea that AI risk extends beyond technical model performance to include governance, accountability and organisational processes.

Human oversight also emerged as a critical element of trustworthy AI. While AI is increasingly capable of analysing vast amounts of information and automating routine tasks, speakers agreed that human judgement remains essential for decisions involving significant financial, legal or ethical consequences.

Rather than removing people from the process, AI is enabling them to focus on higher-value activities such as interpreting complex information, exercising judgement and taking accountability for outcomes.

“An agent can identify a gap for you, but it cannot tell you whether this gap leads to something material or whether a decision needs to be changed. Accountability has to sit with people who have the judgment and regulatory understanding. A machine cannot be accountable,” said Das.

“We have the verifiable, repeatable activities for AI, and the accountable, judgment‑based activities for the human. The boundary between these two is exactly where risk leaders should be placing their attention right now,” she added.

Whether helping clinicians support patients, financial institutions combat fraud or manufacturers simulate future scenarios, the examples shared throughout the conference demonstrated that trust underpins successful AI initiatives. Organisations that invest in governance, transparency and responsible deployment will be better placed to scale AI and realise its long-term value.

“Success won’t belong to the organisations that simply adopt the most AI. It will belong to those who apply AI intelligently, responsibly and with purpose. And that is why trusted AI matters: to give you the confidence that your AI is transparent, that it’s governed, and that it’s delivering value,” said Harris.

Image credit: iStock.com/metamorworks