CFOs Partnering with Generative AI: From Routine Tasks to Strategic Impact

Automating routine work to free strategic time

Generative AI is beginning to change how finance teams spend their time. By handling repetitive, time-consuming tasks, large language models (LLMs) and related tools can free CFOs and their teams to focus on strategic priorities: advising on financial strategy, scenario planning, and responding to geopolitical and market uncertainty.

Andrew W. Lo, Charles E. and Susan T. Harris professor and director of the Laboratory for Financial Engineering at the MIT Sloan School of Management, highlights the practical value: “LLMs can’t replace the CFO by any means, but they can take a lot of the drudgery out of the role by providing first drafts of documents that summarize key issues and outline strategic priorities.” In other words, generative AI can accelerate document preparation and initial analysis so finance leaders can spend more time interpreting results and advising the business.

Practical use cases across the finance function

Adoption is already visible in multiple finance areas. CFOs and finance teams are using generative AI to draft quarterly reports, craft investor communications, and produce concise strategic summaries. Treasury functions are experimenting with cash, revenue, and liquidity forecasting and management. The technology is also used to automate contract processing and assist with investment analysis.

Deloitte’s 2024 State of Generative AI in the Enterprise survey found that about one-fifth (19%) of finance organizations have adopted generative AI in the finance function, signaling early but growing interest.

Forecasting challenges and mathematical limits

Despite promising applications, LLMs have limitations—particularly for tasks that require rigorous mathematical forecasting. The probabilistic and pattern-based nature of many generative models can make precise numerical forecasting and complex quantitative modeling harder to rely on without careful validation and domain-specific augmentation.

That said, combining LLMs with domain models, structured data pipelines, and human oversight can mitigate risks and help teams extract practical value even in forecasting workflows.

Adoption, ROI, and the path forward

Some organizations are already investing despite mixed early returns. Deloitte reports that returns on generative AI investments in finance have been about eight percentage points below expectations so far for surveyed organizations. Still, momentum continues: a Deloitte fourth-quarter 2024 North American CFO Signals survey found 46% of CFOs expect deployment or spending on generative AI in finance to increase in the next 12 months.

Respondents cite cost control through self-service and automation and the ability to reallocate staff to higher-value tasks among the top benefits.

Robyn Peters, principal in finance transformation at Deloitte Consulting LLP, notes that AI can help bring the customer-centric experiences seen in retail and hospitality into finance: “Companies have used AI on the customer-facing side of the house for a long time, but in finance, employees are still creating documents and presentations and emailing them around. Largely, the human-centric experience that customers expect from brands in retail, transportation, and hospitality haven’t been pulled through to the finance organization. And there’s no reason we cannot do that—and, in fact, AI makes it a lot easier to do.”

What CFOs should consider now

CFOs who delay engagement with the technology risk falling behind more agile competitors who are actively experimenting. Finance leaders should:

• Identify high-volume, low-value tasks (report drafting, standard investor updates, contract extraction) that can be automated or accelerated with generative AI.
• Pilot narrow, well-scoped use cases with clear success metrics and human review loops.
• Combine LLM outputs with structured data models and validation for forecasting and other quantitative tasks.
• Invest in upskilling teams so future finance professionals—already familiar with these tools—can work effectively with AI collaborators.

Download the report.

This content was produced by Insights, the custom content arm of MIT Technology Review. It was not written by MIT Technology Review’s editorial staff. It was researched, designed, and written by human writers, editors, analysts, and illustrators. AI tools that may have been used were limited to secondary production processes that passed thorough human review.