Michael Blanding
Artificial intelligence (AI) arrived slowly and then suddenly. As generative AI was rapidly introduced into the workplace, some people reflexively resisted its intrusion while others grappled with new ethical questions that appeared overnight. “Any time a new technology emerges, there is an opportunity to do good, but also the potential to create problems,” says Dean Scott Beardsley.
“Leadership is needed at many levels: from management, engineers, shareholders and even users of the technology. They need to understand exactly what they're dealing with.” Beardsley has just finished a sabbatical from the Uehiro Centre for Practical Ethics at the University of Oxford, where he is researching the relationship between AI, ethics and human well-being.
Professor Yael Grushka Cockayne
The first lesson for those aspiring to lead in an AI-driven world is to confront the reality that AI is here to stay, says Yael Grushka-Cockin, a professor who helped organize a conference on AI leadership at Darden University's Northern Virginia campus last winter.
“There's been a lot of talk about embracing AI and learning to engage with it,” she says. “If you close yourself off and say, 'We're never going to use AI,' that's irresponsible, because you haven't equipped yourself to understand where the dangers are.” The best thing leaders can do is educate themselves about the technology and show they know what they're talking about. “That will earn you credibility and trust in the eyes of your customers, your employees and your investors,” Grushka-Koczyn says. “At the same time, it will allow you to put boundaries and limits on AI.”
aI Str NoOnetheme: and otherslike TItalian HittiyeahgIceBellG MoAbout Effiecient pictureengine
From a strategic perspective, responsible leadership isn't just about ethics, says Professor Mike Lennox, author of a new book. Strategy for the Digital Age: Mastering Digital Transformation (Stanford Business Books, 2023). AI can transform businesses in many ways, from streamlining operations to fundamentally changing how we interact with our customers. For example, companies like Spotify use AI to continuously measure users' listening habits, their immediate social networks, and other similar users to provide them with custom music content. “Companies used to interact with customers occasionally, maybe at a point of sale. Now they're in constant contact with their customers,” Lennox says. This gives rise to new business models and allows companies to transform from selling products to managing networks and platforms.
Professor Mike Lennox.
But as exciting as that is, Lennox cautions that for leaders to embrace AI responsibly, they need to think seriously about the big picture. “Instead of falling into the trap of just adopting AI right away, you need to think about why you're adopting AI, how it can help you, and what you want it to do,” Lennox says. “If you apply AI without thinking strategically, it's like trying to improve the efficiency of the Titanic's engines; you'll just hit the iceberg faster.”
Lennox suggests leaders think about what their company's unique value is to their customers and whether they can better deliver that value with advanced AI capabilities. It also means thinking about whether to “build, buy, or join,” in Lennox's alliterative words: do you create the model yourself, buy an existing AI model, or partner with an AI company to create the model together?
teethsPoncibLu Mingdset: MaAnagiyeahG aFrom me to Be IOnens to a Become familiaryeahEfficient pictureand
It's important to remember that AI is “a means to an end, not an end in itself,” says Ben Reiner (MBA '19), an adjunct lecturer at Darden who teaches the course “Technology and Ethics.” The development of AI creates dangerous incentives to cut corners at the expense of customers, says Reiner, who is also marketing lead for SmartNews, a global news aggregator that uses AI to personalize news for users. “There's often a desire to pursue short-term financial gain without considering the long-term societal impacts and other systemic issues.”
The power of technology, combined with a lax regulatory framework in the United States, where most of the big AI companies are based, may tempt companies to aggregate data for models that could produce better products and services, but also violate user privacy and copyright laws. When dealing with large-scale language models such as ChatGPT, managers may overlook underlying data biases, which could result in models producing outcomes that negatively impact marginalized communities.
“On one side you have people saying, 'Faster, faster, faster. This can really solve some of the world's biggest problems,' and on the other side you have people saying, 'We're at risk of getting out of control. We need to have limits and guardrails so we don't harm people.'”
Scott Beardsley, Dean of the Darden School
While companies can implement safeguards in the way they train their models, they can't completely overcome all bias due to the simple fact that the internet itself reflects the biases of society at large, Reiner said. “We can never remove all bias from these machines,” Reiner said. “But it is incumbent on generative AI companies to understand the models they're deploying so that they can be used in situations where they don't cause unintended harm or perpetuate inequality.”
Lecturer: Ben Reiner (MBA '19)
To navigate the pitfalls of AI, leaders need a clear ethical framework that goes beyond existing regulatory and legal frameworks. “Regulation is the beginning of the conversation, not the end of the conversation,” Reiner said. “Frankly, it’s kind of sociopathic to think that you have the right to do whatever you want unless it’s clearly written into the law.”
For example, consider a theoretical example in which a company is rolling out a new AI model that can identify depression in users, but the product manager in charge notices that the model tends to diagnose a disproportionate number of false positives in women.
Managers might roll out the product anyway and grab market share, hoping they can fix the app before it damages the company's reputation, but truly responsible leadership means a guiding light that sees beyond the financials.
“In my classes, I teach my students to ask themselves, 'If I launch a product now, what is the worst that could happen to various stakeholders, and how would I handle that?'” Reiner says. “In this case, you can revise your goal to say, 'My goal is not just to maximize profits in the short term, but to find opportunities to do so without these false outcomes.'”
circleho is ResponsorbTiltd whatOnet A THey ResResponsible for?
Beardsley says with any new technology, there's a race to deploy it quickly, before it's been fully tested. “On one side you have people saying, 'Faster, faster. This could really solve some of the world's biggest problems,'” he says. “On the other side you have people saying, 'This could get out of control. We need to have some limitations and guardrails so we don't harm people.'” To find the right balance, he says, leaders need to ask themselves up front who is responsible for what, both in terms of immediate impacts and potential unforeseen consequences down the road.
When outcomes are uncertain, companies can roll out new applications on a small scale by conducting pilots or beta tests to minimize the impact if something goes wrong, he adds. University researchers can play a key role in this process by testing AI applications before they are deployed at scale, Grushka-Cockin says, especially in high-risk sectors like health, education and finance. “Even though it feels like the world is rushing at a million miles per hour, traditional, old-fashioned experimentation in a controlled environment is still valid and useful.”
Darden is poised to move headfirst into the field of AI ethics and leadership thanks to a gift of more than $100 million to Darden's Artificial Intelligence Initiative — the largest in the school's history — from David LaCross (MBA '78) and his wife, Kathleen LaCross. The money will be used to support faculty research, develop courses, write cases and host forums to explore the challenges and opportunities of AI in business. “AI is a big business in its own right and a productivity and intelligence tool in many existing industries, including drug discovery, technology, healthcare and education,” Beardsley says. “AI is very relevant to learning and to teaching the next generation what it means to be a responsible leader of the future.”
This new donation strengthens Darden's reputation as a leader in business ethics, with many faculty members widely recognized as thought leaders researching the impact of AI and other disruptive technologies.
Ultimately, Darden's greatest impact may be educating students about the thorny issues that come with the technology and establishing the school as a trusted place for companies to recruit the next generation of responsible AI leaders at all levels of the field.
“Our case method encourages students to be leaders and raise their hands in every situation they encounter in business,” Reiner says. “The question is, how do you use the power that you have?”
