---
id: "claim-cfo-evolution"
type: "claim"
source_timestamps: ["\\\"§ Same Same", "but Different\\\""]
tags: ["cfo", "finance", "skills-shift"]
related: ["entity-russell-reynolds", "quote-chro-architecting-systems", "prereq-traditional-c-suite-functions", "concept-talent-systems-architecture"]
confidence: "high"
testable: true
speakers: ["Tomas Chamorro-Premuzic"]
sources: ["governance"]
sourceVaultSlug: "hbr-seg-governance"
originDay: 7
articleStem: "hbr-sig-56-csuite-board-reshaped-ai"
sourceUrl: "https://hbr.org/2026/06/how-c-suite-and-board-roles-are-being-reshaped-around-ai"
sourceTitle: "How C-Suite and Board Roles Are Being Reshaped Around AI"
---
# CFO competencies have shifted from accounting to data science and AI

Based on a dataset of **5,000 open executive roles analyzed between 2019 and 2025** (from [[entity-russell-reynolds|Russell Reynolds Associates]]), the CFO profile has fundamentally changed:

- **Trending up:** data analytics, AI/ML, cloud computing, digital marketing
- **Trending down:** technical accounting, auditing, financial regulations

The modern CFO must **augment their judgment with data and AI**, moving *from reporting to predicting* and *from technical expert to data-driven strategist*. This parallels the CHRO shift captured in [[quote-chro-architecting-systems]] and the [[concept-talent-systems-architecture]]. Appreciating the magnitude requires [[prereq-traditional-c-suite-functions]].

**Confidence: high · testable.**

**External validation (enrichment).** Strongly supported by executive-search and Big-4/consulting trend reports: modern CFOs are expected to be 'strategic, data-driven value creators,' with rising demand for advanced analytics, scenario modeling, and technology enablement. IBM identifies finance as one of five core areas being redesigned for AI. *Caveat:* 'technical accounting, auditing, and regulations trending down' should be read as *relative emphasis* in job specs, not that these skills are unimportant — regulatory complexity keeps technical competence critical. The specific 5,000-role, 2019–2025 dataset is Russell Reynolds' proprietary data; plausible and consistent with external trends but not independently re-quantifiable from open data.