Michael J. Johnson, CFO, CPA, MS

CFO Magazine 

August 2023

BlackINKCFO.com

The dawn of generative AI technologies like ChatGPT and Google Bard coupled with other finance centric Large Language Models (LLMs) and Private Language Models (PLMs) offers an inflection point for the Office of CFO and the broader finance, tax and treasury functions across global enterprises. At a time when the role of CFO is evolving from number-crunching to strategic decision-making FP&A, CFOs must cultivate a nuanced understanding of how generative AI could serve as a pivotal ally. But to move from buzz to hype then on to reality, executives must differentiate between short-term applications and transformative possibilities, while mitigating inherent challenges and risk that it may pose.

Current and Near-Term Applications: Augment, Don’t Replace

Generative AI is primarily known for text and image generation capabilities, making it suitable for tasks that involve a heavy textual component. Whether it's drafting contracts, preparing investor relations communication, or assisting in credit reviews, these tools offer a preliminary layer of automation that complements human expertise. Their prowess in dealing with numerical data, although evolving, is currently limited. Therefore, generative AI should be looked at as an augmentation tool, enhancing existing processes rather than replacing them.

The Future: Co-Piloting the Finance Function

As generative AI matures, we can anticipate its integration with traditional AI tools that excel at numerical analysis. Imagine a scenario where a machine learning model forecasts revenue, and generative AI takes over to explain the variances and even suggest future business strategies. This symbiotic relationship between different forms of AI can empower finance professionals with enriched, actionable insights, essentially acting as an "AI-driven co-pilot" for finance executives.

Transformative Impact: Beyond Efficiency Gains

1. Core Processes: Tasks such as invoice processing, general-ledger reviews, and contract drafting will not only become more efficient but also more effective. As the technology matures, the range of applications will expand, seamlessly integrating with current processes to boost overall operational efficiency.

2. Business Partnering: Generative AI can serve as a potent tool for financial planning and analysis (FP&A), offering insights into various scenarios during budget cycles and delivering faster, more comprehensive business intelligence.

3. Risk Management: From flagging potential non-compliance to predicting possible fraud, generative AI could go beyond detection to offer explanatory and predictive insights, providing a new layer of risk mitigation.

Challenges and Countermeasures

1. Data Accuracy: The journey from GPT-3 to GPT-4 shows promise in accuracy improvements, but there's still a gap. One workaround is to validate generative AI output with human expertise or external tools.

2. Data Security: Using public cloud environments for training or deployment can pose risks. Enterprises may need to consider private cloud or on-premises solutions for sensitive data.

3. Governance and Validation: Currently, there is no standardized governance model for the validation of generative AI output. Companies will have to build internal governance mechanisms to validate and trust the output.

4. “Challenge of Hallucinations within Finance”: Generative AI may produce incorrect yet convincing output. Rigorous testing and human validation layers can mitigate this risk.

A Proactive Blueprint for CFOs

1. Start with Proofs of Concept: For initial applications like investor relations or contract drafting, starting small can offer insights into generative AI’s potential and limitations.

2. Upskill and Reskill Your Team: Identify internal skill gaps and invest in training programs focused on AI literacy and data analytics.

3. Partner with IT: Collaborate to resolve data security issues and prioritize investment in AI technologies.

4. Champion the Technology: Advocate for generative AI applications across different departments and lead by example to promote its benefits.

In a rapidly evolving landscape, inertia is the CFO’s greatest enemy. The technology is advancing; the question is, are you? By understanding the applications, limitations, and transformative potential of generative AI, finance leaders can position their organizations at the forefront of this technological paradigm shift and drive significant enterprise value.

Jump in the while the water is still warm!