How can CFOs navigate GenAI?
Swaminathan Nagarajan
Digital Consulting | Teaching | Career Counselling & Coaching
Gen AI, or Generative Artificial Intelligence, is poised to reshape businesses, and CFOs stand at the forefront of this transformation. What is their role? The CFOs can strategically approach Gen AI adoption at both the enterprise and finance function levels, offering actionable insights to ascend the learning curve rapidly.
The business landscape is constantly evolving due to technology, often bringing about significant changes, and the pace of this transformation is accelerating. Now, a disruptive force is emerging in the form of Generative AI, or gen AI, which has the potential to reshape various industries It's important to note that not all businesses will be impacted in the same way or at the same time. However, regardless of industry or location, gen AI presents substantial opportunities for creating significant value.
Creating value, however, doesn't happen spontaneously. The Chief Financial Officer (CFO) plays a crucial role in allocating resources at the enterprise level, and this allocation needs to be rapid, decisive, and skewed towards projects that promise the highest value generation, irrespective of whether they are driven by gen AI. Similarly, when leading the finance function, the CFO cannot implement gen AI across the entire organization simultaneously. Instead, CFOs should carefully select a small number of use cases that have the potential to deliver the most meaningful impact within the finance function. In this article, we will explore how CFOs can effectively approach the adoption of gen AI on a company-wide scale, prioritize specific use cases within the finance function, and quickly adapt to the learning curve associated with gen AI.
Leveraging GenAI for enterprise value creation
The primary action that CFOs must undertake is the identification of the most substantial opportunities for value creation and ensuring that these opportunities receive the necessary financial and resource support. While Gen AI represents a potentially revolutionary technology, it does not alter the fundamental principles of finance and economics. A company must consistently generate a return that exceeds its cost of capital.
Furthermore, a company's capital, or its access to additional capital, is finite, leading to competition among various projects. To maximize value creation, CFOs must prioritize the 20 to 30 most value-enhancing projects, irrespective of whether they are related to AI. The Pareto principle, where a small number of opportunities generate the majority of cash flows over the coming decade, consistently holds true. It is crucial for CFOs not to neglect the highest-value initiatives simply because a competing project has the label of "gen AI" attached to it. Ultimately, shareholders should not bear a premium for Gen AI, but CFOs must also be vigilant about existential threats to the company's businesses and be clear about the critical factors for generating and sustaining higher cash flows.
When an opportunity directly addresses or significantly relies on Gen AI, CFOs should not dismiss it due to a lack of understanding or imagination regarding the technology's potential value. Capital allocation decisions are often not binary choices, as incorporating Gen AI can have a more substantial impact on critical business aspects, whether they are revenue generators, margin expanders, or factors that span both revenues and costs.
For example, Microsoft has made significant investments in Gen AI, creating tools like Copilot within Microsoft 365 that provide real-time suggestions to improve documents and presentations. Even traditional non-technology companies like Morgan Stanley's Wealth Management division have made remarkable progress by using Gen AI to provide relevant content and insights to financial advisers.
A world-class CFO ensures that Gen AI initiatives receive adequate funding, debunking the misconception that the CFO's role is to wait and see or be a naysayer. Capital should be actively deployed to support profitable growth. The best CFOs embrace innovation, continuously learning about new technologies, and preparing businesses for rapid technological advancements.
While CFOs should exercise caution, they should actively seek information about opportunities and threats associated with Gen AI. They should collaborate with senior colleagues to define the organization's risk appetite and establish clear risk management guidelines well in advance of the test-and-learn stage of a project.
Some CFOs may perceive championing visionary innovation as contrary to their role as "numbers people." However, they must take on this responsibility because groundbreaking growth often requires more than incremental changes. CFOs can use their relationships with functional and business unit leaders to encourage exploration of Gen AI opportunities, upskilling their team members to foster better cross-organizational understanding and innovation project support. Furthermore, they should remain engaged with innovation efforts continuously, rather than just during periodic reviews or when facing challenges.
New Technology/New Risks
The Chief Financial Officer (CFO) often serves as a company's de facto chief risk officer. Even when a company has a dedicated risk team, as is common in financial institutions, CFOs remain integral in identifying and addressing risks. The advent of Generative AI, or Gen AI, introduces a host of new challenges and potential risks. The adage "to err is human; to really foul things up requires a computer" has gained new relevance in this context.
1. Potential for Egregious Errors: Even the most advanced Gen AI tools are not infallible. These systems, which aim to replicate human preferences and decisions, can sometimes generate results that are convincingly presented but fundamentally nonsensical. For instance, a leading Gen AI platform once produced a legal brief that appeared sound but contained citations from imaginary court cases and quotes supposedly uttered by judges, fabricated by the AI model.
2. Financial Reporting Errors: Gen AI models may produce financial reports that appear flawless at first glance but are factually incorrect. The line items may not correspond to the company, and the mathematical calculations might appear accurate but do not add up correctly. What initially seems like a legitimate financial report can be entirely disconnected from reality.
3. Intellectual Property and Data Rights: The source of Gen AI models often relies on intellectual property rights, not only for presenting information but also for processing data to train the AI. This raises legal concerns regarding the usage of proprietary data, making Gen AI particularly relevant for internal finance functions where company data is often proprietary.
4. Privacy Concerns: Utilizing third-party Gen AI models may expose a company to privacy breaches, potentially revealing confidential or market-sensitive information to external systems.
5. Model Bias: Gen AI models can inherit biases from the data they are trained on, potentially leading to skewed or unfair outcomes in decision-making processes.
6. Tail Event Errors: Overreliance on Gen AI without human oversight can result in unforeseen errors, especially during extreme or rare events, highlighting the importance of human intervention for stress-testing the solutions generated by AI.
7. Reduced Preparedness: A lack of understanding of the underlying analyses or data can diminish the ability of finance teams to critically evaluate the reasonableness of Gen AI outputs. It's crucial to remember that Gen AI is designed to enhance human productivity, not to replace human judgment entirely.
In navigating the world of Gen AI, CFOs play a pivotal role in managing these risks. They must strike a balance between harnessing the benefits of Gen AI for enhanced efficiency while remaining vigilant in identifying and mitigating potential pitfalls. Gen AI should complement human decision-making, not supplant it, ensuring that the finance function continues to operate effectively and responsibly in the face of this transformative technology.
领英推荐
GenAI and Finance Function
Many finance functions are set to integrate Generative AI, or Gen AI, into their operations, marking it as one of the essential tools in their arsenal. This technological advancement holds the potential to significantly save time and resources, making it a compelling reason for adoption. Consequently, it's likely that most, if not all, finance functions within large enterprises will incorporate Gen AI into their workflows within the next three to five years. To contextualize Gen AI's role, we can consider it as the "third wave" of digital transformation.
Getting Started in the Finance Function
CFOs need not be software engineers or AI experts to begin integrating Gen AI into their operations. Their initial step should involve experimenting with the technology to understand its capabilities and limitations. Solutions like OpenAI's ChatGPT and McKinsey's Lilli are readily available online. CFOs can start by uploading publicly available earnings call transcripts from competitors and requesting the AI tool to generate the five most frequently asked questions along with suggested answers. Alternatively, they can upload their own company's financials and those of competitors, asking the Gen AI solution to adopt the perspective of an activist investor, identifying key performance elements that an activist might scrutinize. Depending on the sophistication of the Gen AI solution, CFOs can also upload invoice and payments data, requesting the creation of visualized charts, including a request for the most critical chart.
Through hands-on experience, CFOs gain a deeper understanding of Gen AI's capabilities and can identify immediate opportunities. We recommend budgeting a nominal amount during the learning phase, not for full-scale AI deployment but to enhance the learning experience for themselves and their teams. However, the focus should be on selecting a few high-impact use cases, ideally two to three, and initiating action rather than attempting to explore every possibility.
It's essential to recognize that Gen AI is not a plug-and-play solution. It doesn't create like a human, have eureka moments, or perform mathematical calculations (which falls under traditional analytical AI). Gen AI functions as a predictive language model, translating unstructured data into content that is appealing to humans. Therefore, rigorous processing and curation of data sets are prerequisites, akin to how data scientists prepare data lakes for advanced analytics and analytical AI applications.
Identifying Use cases
Gen AI presents a diverse array of use cases, from customizable interactive charts through natural-language queries to digital performance management, first drafts of external reporting, and working capital management with features like support bots for collections and payments. Gen AI has the potential to revolutionize finance functions across three pivotal dimensions:
1. Automation: Gen AI can effectively automate laborious tasks, such as generating initial drafts of presentations. By taking over these repetitive duties, it liberates valuable human resources for more strategic activities.
2. Augmentation: Gen AI serves as a force multiplier, enhancing human productivity and efficiency. It excels at gathering and synthesizing multiple pieces of information to create coherent narratives, thus aiding finance professionals in their decision-making processes.
3. Acceleration: Gen AI is adept at extracting and indexing knowledge, resulting in a notable reduction in financial reporting cycles and an accelerated pace of innovation. This accelerative capability empowers CFOs to proactively manage performance and support critical business decisions.
A high-performing finance function recognizes the specific use cases where Gen AI can have the most significant and feasible impact, thereby optimizing its operations and contributing to organizational success
In the following chart, one can see some very common use cases being considered/built by the finance function.
1. Synthesis of Information: Gen AI excels in synthesizing data and can create customizable interactive charts through natural-language queries. For instance, it can include solutions like a general Q&A chatbot, a chart generation tool capable of producing charts within seconds based on textual prompts or code descriptions, and a visualization tool that customizes charts using existing code while ensuring its accuracy.
2. Digital Performance Management: Gen AI plays a crucial role in answering performance-related queries, synthesizing status and scenarios, identifying drivers behind budget variances, and proposing resolutions. This self-serve solution is designed to be user-friendly for business users, fostering more effective performance management dialogues.
3. First Drafts of External Reporting: Gen AI can streamline the process of preparing advanced first drafts for various reports, including securities filings and sustainability reports. It can also run queries on current regulations and standards, ensuring that reports align with the latest requirements and guidelines.
4. Working Capital Management: With features like an ever-ready support bot for facilitating collections and payments, Gen AI optimizes working capital management. It provides an always-updated assessment of customer payment history, even enabling real-time adjustments to customer credit limits based on specific activity and market events.
To Sum up: