AI, Trust, and the Expanding Role of Finance: A Sage Future Special
At the Sage Future Conference, three Sage leaders—CTO Aaron Harris, SVP John Fasoli, and SVP Julie Adams—discuss how AI is transforming finance operations, emphasizing that trust, transparency, and accountability are the critical gatekeepers to broader AI adoption. Harris argues that the future of finance lies in AI empowering strategic decision-making rather than just automating repetitive tasks. Fasoli and Adams ground these ideas in practical applications across finance teams and the construction sector respectively.
Summary
The episode features three interviews conducted at the Sage Future Conference in San Francisco, each offering a different vantage point on AI's role in reshaping finance.
Aaron Harris, Sage's CTO, opens with a macro-level view of where AI is headed. He argues that CFOs face an 'emotional leap' in moving from AI as an assistive tool to AI as an autonomous agent—and that this leap is gated entirely by trust. Harris introduces the concept of 'black box versus glass box' AI, insisting that transparency and explainability are prerequisites for adoption. He predicts that within five years, the conversation will shift from AI automating routine tasks like invoicing and reconciliation to AI enabling consequential strategic decisions—such as market entry, acquisitions, and product launches. He also warns that unlike features or speed, trust cannot be hyped or accelerated, putting established vendors like Sage at an inherent advantage over newer entrants. On governance, Harris argues that AI agents must operate within the same permission and access frameworks used for human workers, but with additional layers of traceability—including capturing the prompt, the initiating actor, and which agents were involved. He distinguishes between areas where AI should recommend versus act autonomously, suggesting that for consequential decisions, AI's role is to empower the decision-maker with full context rather than to render a verdict.
John Fasoli, a senior leader focused on finance automation, brings the discussion to ground level. He emphasizes that in finance, '99% right is still wrong,' and argues that AI tools must be built with inspection and validation mechanisms that allow users to verify outputs before trusting them. He describes how high-performing finance teams are distinguished not by recklessness or passivity, but by aggressive experimentation in controlled environments—starting small, validating against historical records, and scaling gradually. Fasoli frames the human-in-the-loop model as essential, particularly for gray-area decisions that require accounting expertise and business judgment. He also argues that as AI eliminates administrative work, finance professionals will increasingly be called upon to handle exceptional situations—either edge cases the AI cannot resolve or high-stakes strategic scenarios requiring human expertise. He sees the finance function becoming more strategic as real-time data access improves decision-making speed and quality.
Julie Adams, who leads Sage's construction and real estate segment, applies these themes to a notoriously technology-laggard industry. She outlines the current challenges facing construction finance leaders: labor shortages, tariff-driven material cost uncertainty, and the critical need for real-time project visibility across disconnected systems. Adams argues that AI has the potential to transform the entire construction value chain—from pre-construction bid optimization and anomaly detection in estimates, to AI-powered drawing management in the field, to labor cost tracking tied directly to project financials. She highlights that Sage is embedding AI across its end-to-end suite rather than offering it as a bolt-on, and is making AI capabilities available even in its on-premises products to meet customers where they are in their digital maturity journey.
Across all three conversations, a consistent theme emerges: the opportunity AI presents is not merely about automation efficiency, but about connecting data, improving real-time visibility, and elevating the strategic role of finance leadership within the enterprise.
Key Insights
- Aaron Harris argues that trust cannot be hyped or accelerated, giving established vendors with decades of market credibility a structural advantage over AI startups that focus on showcasing dramatic capabilities rather than building confidence.
- Harris predicts that within five years, finance conversations will shift away from AI automating invoices and reconciliations toward AI helping CFOs make consequential decisions—such as whether to acquire a company, enter a new market, or launch a product.
- Harris contends that for truly consequential business decisions, AI's role should not be to recommend an answer but to equip the decision-maker with the broadest possible view of data, risks, and context so they can act confidently.
- Harris observes that CFOs are increasingly being chosen to lead digital transformation initiatives because their professional orientation around trust and confidence aligns with what technology-driven transformation requires, expanding their role beyond financial performance.
- John Fasoli argues that high-performing finance teams are distinguished by aggressive curiosity and willingness to experiment in controlled environments—validating AI outputs against historical records—rather than either avoiding AI or deploying it without guardrails.
- Fasoli claims that AI agents built on Sage's platform are subject to the same permission and control mechanisms applied to human team members, because unconstrained agents can have significant unintended consequences on the fidelity of underlying business data.
- Fasoli predicts that AI will eliminate virtually all administrative and repetitive work in finance, with humans increasingly reserved for exceptional cases—either situations AI cannot resolve or high-stakes strategic decisions requiring expertise and judgment.
- Julie Adams argues that AI's greatest value in construction is not in one discrete function but across the entire project lifecycle—from bid accuracy and pre-construction estimate anomaly detection to field drawing management and labor cost alignment—and that this end-to-end integration is where the technology's true value is realized.
Topics
Full transcript available for MurmurCast members
Sign Up to Access