Q&A with Alexander Bant, vice president for the Gartner Finance practice
Generative AI will play a significant role for CFOs who plan to implement an autonomous finance function, but amid lofty claims of its potential society-transforming potential, details of specific use cases are often lacking.
Experts at the Gartner CFO & Finance Executive Conference, taking place September 18-19 in London, are discussing the practicalities of implementing generative AI in a corporate finance function. Alexander Bant, vice president for the Gartner Finance practice and conference chair, expressed his views on what use cases generative AI has in finance, and what forces are driving adoption.
Q: What are the top five use cases for generative AI in the finance function?
A: When we look at some of the more publicized uses of generative AI, such as mimicking well-known authors, it might be hard to see how this can apply to a finance function. But an internally managed generative AI, as opposed to public-facing tools such as OpenAI’s ChatGPT and Google Bard, trained on corporate data has the potential to perform important tasks within finance. The top five that CFOs are exploring are:
1. Contract & Document Review
Generative AI can scan contracts for errors and specific terms. Additional algorithms allow users to ask questions using natural language to get answers about terms and provisions. The same algorithms are used to summarize and categorize documents for sorting, review, and retrieval.
2. Financial & Management Reporting Draft Creation
Generative AI can compose first drafts of management analysis and discussion talking points, as well as financial footnotes that finance teams evaluate and refine.
3. Policy Interpretation
Generative AI can review large collections of existing financial policies, like T&E policies, and provide initial recommendations for how those policies could be applied for finance teams to evaluate and refine.
4. Coding Assistance
Generative AI can translate code from older coding languages, like COBOL, into more modern programming languages, like SQL, KnowledgeSQL, and Python.
5. Forecast & Budget Variance Explanation
Generative AI can provide explanations of forecast and budget variances for FP&A teams to use in business reviews, as well as further synthesize those trends and insights for executive and Board consumption.
Q: What forces are driving generative AI adoption in corporate finance?
A: There are several, but arguably the most direct driver is that boards and C-suites are impressed with the potential for innovative generative AI use cases to drive growth and profitability. If implementing generative AI can make the finance function more agile, and able to provide higher quality analysis and better strategic support to the business, it could be a significant source of competitive advantage for leading adopters.
Further, investors expect new sources of growth and productivity, and ultimately better margins. This is placing pressure on leaders to not get left behind in the AI race.
While many employees have concerns about job losses due to AI, if they begin to see the technology as an essential part of doing their work, they will be more likely to quit organizations where they aren’t able to fully leverage generative AI.
Lastly, regulators expect organizations and their leaders to comply with responsible generative AI regulations, and this will likely lead to adoption in a formalized and controlled manner rather than leaving employees to experiment with the technology.
Q: Where are finance functions on the journey of adopting generative AI?
A: Naturally, leadership teams don’t want to fall behind peers; however, as the chief steward for an organization’s financial health, CFO’s must balance the risks and rewards of tools like generative AI.
Fewer than 10% of finance organizations talking to Gartner are using generative AI in production currently. Within a few years, however, the distribution is likely to follow the curve of enterprise adoption.
Almost all enterprises have taken some steps in reaction to generative AI’s emergence. Typically, the first move is to protect data assets and IP, by ensuring that usage is in a considered and controlled manner.
Most enterprises today, around 60%, are assessing potential use cases and associated risks without making any significant implementation.
Around 30% of enterprises have identified a few specific use cases and are testing generative AI in an iterative process, while the leading 10% are seeing gains from applying the technology extensively and are looking to expand its use.