Why BioPharma C-Suites Can’t Delegate Accountable AI to Their Chief Legal Officers

Kevin Werbach, Professor of Legal Studies and Business Ethics at the Wharton School.

Kevin Werbach is a professor of Legal Studies and Business Ethics at the Wharton School, University of Pennsylvania, and formerly Counsel for New Technology Policy at the US Federal Communications Commission. Professor Werbach served on the Obama Administration’s Presidential Transition Team and helped develop the United States’ approach to internet policy at the FCC during the Clinton Administration. He is the host of the podcast The Road to Accountable AI.

He also created one of the most successful massive open online courses, with nearly a half-million enrollments. Werbach earned his BA from the University of California at Berkeley. At Harvard Law School, he was the publishing editor of the Harvard Law Review.

Q. With NVIDIA’s BioNeMo leveraging generative AI to accelerate the drug discovery process, several leading pharma/biotech firms are aggressively tapping into the platform to analyze vast datasets with unprecedented speed and precision so that they can identify potential drug candidates more efficiently.¹ Still given the potential risks of these emerging technologies, how can their chief legal officers provide guidance on building an AI daily operating model, which is responsible, safe, and trustworthy at their respective firms?

A. Werbach. Accountable AI is not only for the purview of chief legal officers (CLOs) but also for their C-suite colleagues and boards. As for the CLOs, while they are on point to help mitigate risk, these leaders need to recognize how the legal departments can’t be just about saying “No” but rather about enabling their peers to understand risks and to move forward with decisions that can support C-suite business goals. At Wharton, we have ten academic departments and already half of them are participating in my Accountable AI program.²

For biopharma firms which have global operations and complex regulatory, compliance, and legal considerations given local country considerations, CLOs should especially provide guidance around data governance. With the vast and only accelerating numbers of new data sources being channeled into their organizations, CLOs should take the lead to break down silos and work constructively with peers like CIOs and business unit executives to address key challenges like data protection, cybersecurity, and other strategic considerations.

Q. While AI has been around for decades, its potential to drive tangible business results has expanded beyond historically targeted areas like sales and marketing. For example, it has been interesting to see how AI has been deployed in the legal ecosystem.

According to a partner at Cravath, Swaine & Moore,³ "We have some pilots underway…in the area of due diligence. Some clients ask us to use AI tools to review very long contracts that would have required an army of lawyers to do it."

Preliminary findings showed how on one occasion, it took hours rather than weeks to review these hundreds of contracts. Thereafter, tests were run to check the accuracy of the reviews and a 95% rate was calculated. Also, the firm had mentioned that one of the advantages of generative AI is how lawyers can engage with the tools in natural language.

Lawyers can simply ask it to show them drafts of M&A deals in a particular type of industry and with certain characteristics. As such, it is a more intuitive interaction and also faster. They can even ask the AI tool to summarize the information or present it in graphical form.”⁴ Given these emerging trends, what do you see on the landscape for Fortune 500s when it comes to deploying AI in an accountable manner?

A. Werbach. Most Fortune 500 firms have been leveraging AI for several years. As for generative AI, it has exploded during the past two years with employees excited about the potential to uncover new innovative ways to leverage it, despite business risks such as hallucinations. But leaders need to recognize the importance of transparency, which refers to how humans can understand and trust the outputs created by these emerging technologies.

By doing so, these C-suite leaders can help address the typical black-box questions that people have around the model’s accuracy, fairness, and transparency. While these efforts do require resources and time, the good news is that forward-thinking companies are making solid progress. For example, at Walmart, they deployed a “Walmart Responsible AI Pledge” and have invested resources so that continuous monitoring can help ensure AI tools align with their corporate values that are in place.

Of equal importance, their executives have implemented pragmatic resources, like their "Ask Sam" tool, which aids their frontline associates to elevate their support of Walmart’s customers. For instance, Walmart’s execution of a conversational AI platform has produced concrete business results saving customers’ and agents’ time; and resulted in customer satisfaction scores increasing by about 38%.⁵ With ever-changing developments like the recent EU AI Act,⁶ the road to accountable AI is a journey and your readers can learn more about potential new pathways from a variety of leaders here.

About the Author

Michael Wong is a Part-time Lecturer for the Wharton Communication Program at the University of Pennsylvania. As an Emeritus Co-President and board member of the Harvard Business School Healthcare Alumni Association as well as a Contributing Writer for the MIT Sloan Career Development Office, Michael’s ideas have been shared in the Harvard Business Review and MIT Sloan Management Review.

References

1. https://investor.nvidia.com/news/press-release-details/2024/NVIDIA-Healthcare-Launches-Generative-AI-Microservices-to-Advance-Drug-Discovery-MedTech-and-Digital-Health/default.aspx

2. https://accountableai.net/

3. https://www.cravath.com/news/cravath-tops-2025-vault-law-100-ranking-of-most-prestigious-law-firms.html

4. https://neosmart.ai/cuatrecasas-cravath-and-arent-fox-schiff-tell-their-experience-using-ai/

5. https://tech.walmart.com/content/walmart-global-tech/en_us/blog/post/three-ways-we-are-using-conversational-ai-at-walmart.html#:~:text=Ask%20Sam%2C%20our%20voice%20assistant,or%20%E2%80%9CWhat's%20my%20schedule%20for

6. https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai