Trust Is the Operating System of the Agentic Enterprise

Trust Is the Operating System of the Agentic Enterprise   In this episode, Sabine VanderLinden is joined by Franklin Manchester, Global Insurance Strategic Advisor at SAS, and Steven Abel, Global Technology Partner and Deputy Global Head of AI & Transformation at Oliver Wyman. Together, they unpack the concept of "trust by design" in the context of agentic enterprises and AI adoption.    The conversation pivots from traditional risk frameworks and compliance-based approaches to trust, to the urgent need for architectural and cultural transformations in which trust is embedded in every system and decision. They explore why organisations often confuse expanding AI tools with genuine readiness for autonomy, discuss why "human in the loop" is no longer sufficient, and offer perspectives on scaling trust, managing risk, and redefining organisational roles.    The trio debates actionable leadership moves for CEOs and boards, the evolving skills for insurance professionals, and how the frontier firm of the future will distinguish itself through intentional trust-building—not just AI deployment.   KEY TAKEAWAYS   Many organisations treat AI trust as a compliance issue, which hinders safe scaling. The fundamental shift involves deploying autonomous decision-makers, making trust by design an architectural and leadership mandate.   We, as an industry, over-invest in AI models and technology while under-investing in people and trust. Simply using more models or data doesn't guarantee higher trust, especially without architectures built for transparency and governance. Franklin noted a disconnect where insurers use AI but lack trustworthy systems, surprisingly favouring newer generative AI over established machine learning.   I question the efficacy of "human in the loop" controls in high-stakes industries, while Steven advocates embedded, infrastructure-level trust solutions. Franklin identified processes as primary failure points, particularly when tacit knowledge is overlooked (citing Cigna's mass claim denials).   The discussion explores the need for new AI risk and governance roles, akin to past actuarial practices. While human-centricity should drive design, scalability is challenging as organisations move toward agentic systems in which humans supervise, rather than directly control, risking brand integrity if governance fails.   For leaders, I urge you to shift focus from technology hype to foundational trust. Steven prioritises "under the water" capabilities, such as risk and regulatory expertise. Franklin recommends three people-centric actions: embracing new skills, breaking data silos, and protecting the brand.   The truly future-ready firm embeds trust into every decision system—a practice rooted in culture, governance, and leadership, not just technology. Scaling AI without trust is merely scaling risk; organisations must engineer trust as a core operating principle.   BEST MOMENTS   "Trust isn't what you say, it is what your system does." — Sabine VanderLinden   "The architecture of these models themselves don’t lend themselves to a high trust environment." — Steven Abel   "We trust generative AI 200% more than machine learning. Which is bonkers to me because machine learning has been around for like 30 years." — Franklin Manchester   “There’s still no more sophisticated sensor than a human being and a more powerful computer than the human brain.” — Franklin Manchester   “Auditability, transparency, and a connection with the human ecosystem and judgment—these things are non-negotiable.” — Steven Abel   "It is clear that the adoption is moving fast, and we need to make sure within regulated industry that we apply trust in everything we do. Otherwise, we are going to shun both customers." — Sabine VanderLinden   ABOUT THE GUEST   Franklin Manchester Prior to joining SAS, Franklin served as a Global Insurance Strategic Advisor at SAS Institute, bringing over 20 years of experience in insurance underwriting and analytics. Known for his deep industry insight and passionate advocacy for trustworthy AI, Franklin is currently focused on linking insurance expertise with AI-driven transformation, highlighting the importance of governance, ethical frameworks, and human-centricity in future-ready companies.