Generative AI Didn’t Write This—and It Wouldn’t Tell You If It Did

By Kris Pederson


Artificial Intelligence Technology Oversight Online Article

With its November release, ChatGPT, a generative artificial intelligence (AI) tool, ushered in a renewed interest in artificial intelligence and related technologies. Generative AI models use a technique that trains AI to generate new things, such as sentences, images, or videos. “ChatGPT” and “AI” are increasingly being seen in headlines and talked about in the boardroom.

One report by Bloomberg noted that references to AI and related terms in earnings calls were up 77 percent by March 1 compared with the first two months of 2022. At least one news organization has made use of AI to draft articles—77 of them between November and January, to be exact. Many curious staff, executives, and directors have logged onto generative AI platforms and thought, “How much of my work could this actually do?” The answer for some is quite a lot.

This should give board members some pause. Over the last several months, the EY Center for Board Matters has engaged with the director community to understand the board’s role with respect to these new technologies. Several themes have risen to the top.

Don’t get distracted by shiny new technologies. One important role the board can play is to keep the management team focused on a coherent longer-term strategy. While ChatGPT and similar generative AI platforms have sparked curiosity, from the board perspective these technologies should be viewed as just one application of AI. Directors should encourage management teams to think in terms of customer or business value that can be created through intelligent automation tools such as AI. Think about not just piloting AI as a technology, but piloting a new way of working—using generative AI to more quickly create engineering designs or the first draft of a new committee charter, for example. Because of this, ChatGPT and related technologies may also disrupt content-focused businesses and functional roles. The board should help the management team address broad business model changes that are likely to take place as these technologies are more fully integrated into the ways in which people work and live.

Get hands-on with the technology. To effectively oversee generative AI and other emerging technologies, directors will need to develop a conceptual understanding of what these technologies do and how they may impact the business. This may come from a variety of sources, such as self-study, education sessions led by a member of the technology team, or sessions from third-party experts. As a director, the key is to know enough about the technology to put it in context and to ask management second- or third-level questions. A proven way to learn new technologies is to use them, so consider—as a learning experience—asking ChatGPT to write a board resolution on new cybersecurity governance or a job description for a new innovation and digital committee chair.  

Recognize that AI will permeate the whole business. Because generative AI is available from any browser, staff can use it to draft memos, policies, and meeting decks. Microsoft 365 is adding a productivity AI bot through Copilot, which will be integrated into its suite of products including Word and PowerPoint. Information technology (IT) teams may use it to write code in a wide array of programming languages. Sales and marketing teams might analyze customer feedback. Customer contact teams can improve voice and text bots that engage with customers. Legal may draft, summarize, and review legal documents through generative AI. A formal AI strategy should identify where AI bets are being placed and how resources are being allocated; the board should have a strong understanding of the key strategic AI implementations and plans. However, given the ease of access to AI, there is likely to be a large amount of shadow AI, akin to shadow IT, which will be used outside formal AI policy.

Understand the known risks and be on the lookout for new risks.  AI’s broad, unprecedented usage likely gives risk-oriented readers some anxiety. AI’s democratization means that an employee can draft a memo with a web-based AI tool and unintentionally share confidential information with a third-party app. AI algorithms can also carry bias, a known risk that has been the focus of AI governance conversations over the past several years. Issues such as ethics, fairness, and trust are not afterthoughts and must be embedded into AI operating models or the business risks losing customer and employee trust. Failure to do so may erode carefully built brands. Additionally, there are unknown risks that will reveal themselves as the adoption of the technology—and its use cases—expands. Directors should weigh an effective AI risk appetite to balance the risk and rewards of generative AI as it matures. Generative AI may present challenges to existing business operations, employees, ethical standards, and liability risks and businesses should be prepared to assess and manage those risks effectively. 

Consider how to incorporate AI in the board meeting. Finally, directors may be well served by looking at AI in relation to board operations themselves. Key elements of the board meeting present opportunities to automate and modernize, which may improve the learning curve and familiarity with generative AI. Though virtual meetings and digital board books have been significant innovations over the past 20 years, most processes and procedures have remained unchanged and primarily manual. Directors may consider working with their firm’s corporate governance team to consider how generative AI may improve the efficiency and effectiveness of board reporting to allow for more rich discussions on the firm’s most important issues.

AI is exciting and now it seems to be everywhere. It is showing up in your browser, likely in your office, and in other areas of your professional and personal life. Companies should be experimenting with where AI applications might best be used with an eye toward creating new business and customer value through an AI strategy. This strategic application of AI is where boards are in the best position to assist management by keeping a focus on the long term. In the meantime, directors themselves may be well served by experimenting with generative AI to see how the technology can draft letters, respond to questions, and write articles such as this one. I promise I wrote this myself without ChatGPT—at least this time.

Kris Pederson
Kris Pederson, NACD.DC, is the Americas leader of the EY Center for Board Matters.