Understanding the ethical and business imperatives of generative AI

But that also means 80,000 said no. 

“The onus is on us as employers,” said Arvind Krishna, chairman and CEO of IBM, in explaining his desire to utilize AI tools to help employees become more productive. “How do we train everyone inside to become a user and use it to improve their job?”

Krishna believes generative AI will greatly impact white-collar jobs, especially those that are easily automated and with tasks that are repetitive. But larger societal challenges emerge as employers continue to increasingly incorporate AI in the workplace: Can everyone make the transition? 

“There is going to be some displacement,” said Krishna. “And the question is, what are the roles for those people who cannot come up the cognitive curve?”

CEOs that joined together for a Fortune virtual event discussing the ethical and business imperatives of generative AI were in agreement that the technology offered great productivity benefits, as well as liabilities, that every industry must weigh as they move to invest in AI. But some of the loftier commentary focused on how the world will also change. 

As population growth slows in Western countries across Europe, Japan, and the United States, employers may need to rely on the productivity jolt from AI to maintain annual gross domestic product growth of at least 3% to help maintain economies. 

“Since the industrial revolution, problem solving has been a human endeavor,” said Ravi Kumar, CEO of Cognizant. “This is an inflection problem for problem solving dedicated to machines, and purposeful problem finding is the new human endeavor.” 

He sees the benefits of AI to propel upward social mobility of the workforce, asserting that AI can help eliminate pain points in the workplace, ultimately making for a better environment.

“I think generative AI will humanize technology and hand over the agency to the end user,” said Kumar.

A recent survey commissioned by Boston Consulting Group with MIT Sloan found that 84% of BCG’s clients were aware of the risks of AI but only 16% thought they had a mature way of responding. Risks can also stem from the elements of AI that impact strategic aspects of work, which could include a screenwriting for a movie or drug discovery. 

“Make sure you have someone who’s responsible for it, and at the top of the organization, don’t delegate it down,” said Sharon Marcil, North America chair and managing director and senior partner at Boston Consulting Group. “And then ensure that this risk protection is embedded in your risk processes, but also your core processes like product development, so that it really becomes a cultural element.” 

There are already efforts underway at many companies to use AI to improve specific functions like customer service and marketing, areas where the development of guardrails isn’t so difficult. But there are a lot of potential liabilities if AI unknowingly infringes on intellectual property. 

Krishna gave the example of a possible client in the payments industry, where developers use AI to create code that might be used for 20 years. But 15 years into using that code, someone may discover that there’s been potential infringement. 

“This is going to get into a set of issues that are far deeper than some other areas,” said Krishna. “And we are getting some questions from our clients.” 

One possible solution? The provider of the technology has to share some of the liability with clients.

Some are in the earlier stages of adoption. IKEA Group CEO Jesper Brodin had been in a training session with 400 leaders about AI earlier in the day of the Fortune virtual event—and has two more sessions planned to explore opportunities for AI and see how it fits into the business. IKEA already uses generative AI in a few ways, but the tech is quickly picking up speed in adoption at retailers.

“Leaders, in general, are not there yet,” said Brodin. “So that’s why we think it’s important to get everybody on board.” 

Blue Cross Blue Shield Association (BCBSA), which provides health insurance to 118 million Americans, is similarly a bit cautious. Kim Keck, president and CEO of BCBSA, has to weigh two ends of the spectrum: the need to have guardrails in place to protect data but also consider innovation. 

“We’re spending a fair amount of time advancing guardrails on the risk side and what that framework could look like,” said Keck. 

But on the innovation side, BCBSA has to consider that it pays 600 billion claims a year. There are a lot of efficiencies to be gained with greater usage of generative AI.

“We are in the early stages of organization adoption,” said Barak Eilam, CEO of Israeli software company NICE. “When I speak to our customers, we are in the phase of AI FOMO: the fear of missing out.” 

Eilam cautions there is a bit of overestimation about how fast and how large the immediate savings can be achieved from generative AI. Where he sees great potential is changing the mindset of workforces and getting talent to become more curious about exploring the tech.

Tom Wilson, chair, president, and CEO of the Allstate Corporation, sees AI as a spectrum. There are large language models that already exist today, machine-based learning that’s getting better, and other aspects of AI that aren’t yet a reality but will be some day. 

“We’re just moving farther along the spectrum,” said Wilson. “It’s a bigger slope in terms of the degree of change than it has been in the past.” 

At Allstate, Wilson is thinking about how to build an AI ecosystem within the insurer. Some areas are obvious—call centers, as an example—and the tech is better developed externally. But the development of AI within Allstate requires a shift in thinking from a strategy supported by technology to a technology-driven strategy. 

“For us, it is more about building the internal capabilities that we can’t get from outside, and then making sure we have the partnerships with the people who are outside, who bring us their best thinking,” said Wilson.