CFOs have fundamentally shifted their approach to artificial intelligence (AI), according to new research from Salesforce, moving from cautious spenders to strategic investors who are betting on AI not just for cost-cutting but as a crucial engine for long-term revenue growth.
A striking 70% of the 261 global CFOs surveyed reported having a conservative AI strategy in 2020. Fast-forward to today, and that number has plummeted to a mere 4%. This rapid transformation highlights a widespread recognition among financial leaders that AI is no longer just an emerging technology but a crucial tool for enhancing efficiency, optimizing operations, and driving critical long-term growth.
CFOs’ fundamental rethinking of tech investment ROI, according to the data, explains this transformation. More than half (61%) of CFOs say AI agents - digital labour capable of performing tasks autonomously, are changing how they evaluate ROI, measuring the success of technology investments beyond traditional metrics to encompass a broader range of business outcomes.
“The introduction of digital labour isn’t just a technical upgrade — it represents a decisive and strategic shift for CFOs,” said Robin Washington, President and Chief Operating and Financial Officer at Salesforce. “With AI agents, we’re not merely transforming business models; we’re fundamentally reshaping the entire scope of the CFO function. This demands a new mindset as we expand beyond financial stewards to also become architects of agentic enterprise value.”
Last year, in fact, 65% of CFOs faced pressure to accelerate tech investment ROI. Today, they recognize the value of AI isn’t just about short-term cost-cutting but also long-term business outcomes like revenue generation, productivity gains, and improved decision-making — things AI agents are uniquely suited to improve.
“The ROI of older technology often depends on immediate, measurable results,” said one CFO survey respondent, “while AI’s returns may accrue over the long term through an ongoing process and new business models.”
Find more insights on the Salesforce Stat library.
Comments