Q. Earlier this year, McKinsey shared research findings of how while organizations have seen usage with gen AI tools surge, from a value capture standpoint, “…these are still early days—few are experiencing meaningful bottom-line impacts.”[1] And this analysis builds upon similar observations from a few years earlier when MIT Sloan Management Review in collaboration with the Harvard Business School and Boston Consulting Group, conducted a comprehensive study on the enterprise-wide deployment of artificial intelligence (AI) at large companies. As that research pointed to, “While almost 85% of executives believed AI would help their companies obtain or sustain a competitive advantage, and three-quarters thought it would enable entry into new businesses, actual deployment lagged far behind expectations.”[2] Given these continued patterns of high expectations and low value realization, what should be done to catalyze true change?
A. (Iavor Bojinov)
The high failure rate of AI projects, estimated to be as high as 80%[3], underscores a critical disconnect. This rate is nearly double that of typical corporate IT projects from a decade ago. Rather than viewing this as a technology problem, senior executives must recognize it as a leadership challenge. They must take accountability for fundamentally rethinking how their organizations integrate and operationalize emerging technologies into daily workflows. As McKinsey recently shared, nearly all companies invest in AI, but apparently just 1 percent believe they are at maturity. And their research findings, “…the biggest barrier to scaling is not employees—who are ready—but leaders, who are not steering fast enough.”[4], reinforces my assertion that the C-suites are accountable.
As for catalyzing true change, based upon my research and interviews with Fortune 500 C-suites, here are the key playbook questions for your readership to consider:
- Strategic alignment: Is the project in line with the organization’s overarching strategy and goals?
- Measurable impact: Can we objectively assess the project’s financial and operational benefits?
- Augment or replace: Will it enhance current human operations or is it replacing an existing manual process?
- Nature of the problem: Is this a problem AI can solve?
- Data availability: Does the organization have access to the necessary data?
- Technological capability and skills: Does the organization have the infrastructure and skill set necessary to build, deploy, and scale up the project?
- Ethical considerations: Have all the ethical implications been fully considered?
While these executives don’t have to be computer scientists, they should have enough of an understanding of the potential benefits and risks associated with AI so as to lead conversations that delve into the above playbook themes and deploy AI at scale. JPMorgan’s pragmatic deployment of generative AI at scale has been documented as a Harvard Business School case study so that students can understand the bank’s strategy and daily tactics. For instance, the bank’s GenAI toolkit, LLM Suite, is on the desktops of more than 200,000 of their employees, half of whom used it several times a day. Connect Coach and several other GenAI applications have been generating enthusiasm by simplifying and accelerating completion of otherwise tedious and time-consuming “no joy” tasks.[5] And a key takeaway for their employees is how AI can help grow the business for the firm and themselves. The company tracks and communicates measurements like Sales Assist gross sales per user, which experienced an approximate 188% growth between 2022 and 2023. These types of secured benefits have supported their aspirations of leveraging AI as a strategic weapon and building the confidence of prospective employees who are evaluating JPMorgan’s business operations. Indeed, the firm is a popular destination for top MBA students, including MIT Sloan where it is typically an employer of MBA and Master of Finance graduates.”[6]
Q. As biopharmas have historically been considered fast followers of emerging technologies when compared to other verticals like financial services given the complex regulations for drug approvals and other constraints, what are your pragmatic playbook ideas for the readership to consider for the rest of 2025?
A. (Iavor Bojinov)
That may have been true historically, but the landscape is shifting rapidly. Proactive biopharma companies like Moderna are demonstrating that AI can be a core driver of strategy, not just an IT add-on. It’s not surprising given their CEO’s famous quote that Moderna is “a technology company that happens to do biology.” Building upon a case study that my HBS colleagues and I wrote on Moderna, which was founded in 2010 and somehow leapfrogged to Fortune’s 2024 list of the World’s Most Admired Companies, joining the “All-Star” top-50 companies[7]; here are three recommendations.
First, educate your workforce on the benefits of AI for building sustainable careers. Since its beginning, Moderna’s C-suite has consistently explained how and why technology can support Moderna’s revenue aspirations and productivity of its employees. During their early years, Moderna sought to eliminate independent local solutions such as Excel and instead have connected systems, including lab equipment connected through the Internet of Things, where data could be entered once and be conveniently available to everyone who needed the information. Moreover, while many consulting firms sing the praises of building technical platforms at their clients, Moderna’s founders built it at scale via their mRNA technology, which helped the firm provide a solution to the world during the COVID-19 crisis. With traditional drug and vaccine development, companies essentially develop one drug at a time, investing significant capital to do so, in a process that often results in only limited knowledge transfer from one drug to another. With mRNA, Moderna’s C-suite identified an opportunity to develop many drugs in parallel by using similar technologies where cross-drug learning was significant. For instance, a particular LNP that could carry an mRNA molecule into certain types of cells could be used to carry other mRNA molecules into those same cells. Likewise, key learnings in manufacturing one mRNA or LNP could transfer to the manufacture of others. With these past technology playbooks executed and positive outcomes communicated to employees, Moderna staff have embraced the firm’s AI education, and it has been reflected in the team’s embracing of the new tools.
Second, empower employees and show them how to use AI tools to solve pain points that come up in their daily lives at work. For instance, Moderna collaborated with Open AI to deploy Moderna’s own instance of ChatGPT, called mChat. With more than 80% internal adoption since its debut, this initiative led to the deployment of ChatGPT Enterprise and its enhanced capabilities such as Advanced Analytics to support their analytical efforts.[8] And for your biopharma employees as well as consultants selling digital offerings to this vertical, consider how Moderna’s usage of AI has helped address the ever painful process of the legal review of documents. Similar to most organizations, Moderna has a small legal team responsible for providing guidance on 6,000+ documents per year, including NDAs and other client-facing reports. Their CLO recognized how custom GPTs could shift many employees to a self-service model which enabled Moderna employees to query a custom AI agent that would answer their questions in real time. Other GPTs have been deployed which enabled Moderna’s legal function to quickly write new contracts and provide guidance on those presented to them. As the CLO shared with us, “We actually have better compliance now than we did before because people are actually using the GPTs to answer their question. It doesn’t require them to bother a human being or wait for an answer.”[9]
Third, move beyond isolated pilots. The goal should be to use AI as a catalyst for redesigning entire work processes. Instead of marginal improvements, challenge your teams to pursue transformational change. This requires a rigorous testing framework. Especially for the Top 10 biopharma firms, invest in the testing for (AB tests, etc.) how AI might be creating value for your employees. In particular, I recognize that CIOs will get an earful from biopharma commercial leaders who complain if some of their field members are not provided access to certain tools which might help some districts achieve their sales quotas. To build a true learning organization, you must apply the same rigor to internal initiatives as you do to external, customer-facing ones. This means implementing structured experiments, such as A/B tests, for employee-facing AI tools. Without this, you are operating on anecdotes, not evidence. Objective data from these experiments is the only way to build a scalable, data-driven framework for determining which initiatives to start, stop, or continue.
My concluding challenge for your readership is to roll up your sleeves and learn how to actually apply AI tools to make you more effective and efficient every day. Sign up for your employer’s hands-on training event with that instructor who is going to demand that you create some custom GPTs that actually make a difference in your daily lives.
And for the aspiring CEOs in your audience, the message is unequivocal: you cannot delegate AI strategy to your CIO. This is not a technology trend; it is a fundamental shift in how value will be created. As the world’s leading business schools now teach, fluency in AI is no longer optional for leadership—it is a core competency. AI is not the latest fad and top business schools recognize this change. MIT Sloan offers several courses and executive programs focused on AI and its business implications, such as “Artificial Intelligence: Implications for Business Strategy” and the “AI Executive Academy”.[10] Wharton recently announced a new MBA major in “Artificial Intelligence for Business,” allowing students to specialize in AI through a selection of courses covering applied machine learning, data science, ethics, and more.[11] And at the Harvard Business School, students must take an AI course, Data Science and AI for Leaders, to graduate.[12] Learning AI is not optional, it’s required.
Bios:
Iavor Bojinov is an associate professor at Harvard Business School. Professor Bojinov’s research interest is at the interface of causal inference, experimental design, and large-scale computing with the overall goal of democratizing statistical methods in order to help firms innovate and grow. Prior to joining Harvard Business School, Professor Bojinov worked as a data scientist leading the causal inference effort within the Applied Research Group at LinkedIn. He holds a Ph.D. and an MA in Statistics from Harvard and a MSci in Mathematics from King’s College London.
Partha Anbil is a Contributing Writer for the MIT Sloan Career Development Office and an alumnus of MIT Sloan. Besides being VP of Programs of the MIT Club of Delaware Valley, Partha is a long-time life sciences consulting industry veteran. He is a Senior Advisor to NextGen Invent Corporation (https://nextgeninvent.com/), an AI, Data Analytics, and digital transformation company. He has held senior leadership roles at IBM, Booz & Company (now PWC Strategy&), IMS Health Management Consulting Group (now IQVIA), and KPMG.
Michael Wong is a Contributing Writer for the MIT Sloan Career Development Office and an Emeritus Co-President and board member of the Harvard Business School Healthcare Alumni Association. Michael is a Part-time Lecturer for the Wharton Communication Program at the University of Pennsylvania and his ideas have been shared in the MIT Sloan Management Review and Harvard Business Review.
[1] Singla, Alex, Sukharevsky, Alexander, Yee, Lareina, Chui, Michael, Hall, Bryce, The state of AI: How organizations are rewiring to capture value, McKinsey, March 12, 2025
[2] Ransbotham, Sam, Kiron, David, Gerbert, Philip, Reeves, Martin, Reshaping Business with Artificial Intelligence, MIT Sloan Management Review, September 6, 2017
[3] Bojinov, Iavor, Keep Your AI Projects on Track, Harvard Business Review, November-December 2023
[4] Mayer, Hannah, Yee, Lareina, Chui, Michael, and Roberts, Roger, Superagency in the workplace: Empowering people to unlock AI’s full potential, McKinsey, January 28, 2025
[5] Bojinov, Iavor I., Lakhani, Karim R., Lane, David, JPMorganChase: Leadership in the Age of GenAI, Harvard Business School, N9-325-066, April 3, 2025
[6] https://cdo.mit.edu/blog/2024/12/10/2024-2025-mba-employment-report/
[7] https://investors.modernatx.com/news/news-details/2024/Moderna-Named-to-Fortunes-List-of-Worlds-Most-Admired-Companies/default.aspx
[8] https://investors.modernatx.com/news/news-details/2024/Moderna-and-OpenAI-Collaborate-To-Advance-mRNA-Medicine/default.aspx
[9] Bojinov, Iavor, Lakhani, Karim, Hildebrandt, Annika, Weber, James, Moderna: Democratizing Artificial Intelligence, Harvard Business School, 2024
[10] https://mitsloan.mit.edu/mba/explore-program/mba-curriculum
[11] https://news.wharton.upenn.edu/press-releases/2025/04/the-wharton-school-introduces-new-undergraduate-concentration-and-mba-major-in-artificial-intelligence-for-business/
[12] https://www.thecrimson.com/article/2025/4/4/hbs-makes-ai-class-required/
