Good morning. Companies show little sign of slowing down their deployment of AI despite rising concerns that the most advanced models pose serious risks.

In recent days, the heads of leading AI developers have called for new limits and controls on advanced AI research, following a series of cybersecurity breaches and high-profile resignations designed to raise alarm bells. As concerns about potential existential risks reach a new level, the discussion of AI within the corporate world remains measured.

The WSJ Technology Council Summit kicked off Monday afternoon in New York City with an informal poll of the audience during a town hall conversation. In a show of hands, only a few people said they feared that AI could kill us all in the next decade, while about half of attendees indicated that they were in support of slowing down AI development at the frontier and prioritizing safety guardrails.

But that slowdown didn’t apply to the deployment of AI within their organizations, where not all models are at the cutting edge and even the most advanced of them tend to be relatively well understood and tested by experience.

“It’s in our hands and it’s up to us to apply [AI] for good, and I think it can do a lot of good for society,” said Vishal Talwar, president, FedEx Dataworks, and chief digital and information officer, FedEx.

Much of ensuing town hall discussion focused on the myriad ways in which companies were operationalizing AI.

Ruchir Swarup, partner and CIO at KKR, addressed the financial complexities of scaling AI across the massive alternative asset manager.

“As we think about technology across all of our portfolios…one of the things that really stand out for us is how we allocate tokens and budgets to each of those strategies,” Swarup said. “Going down deeper into agent economics basically means understanding what types of tokens and…how much cost is getting attributed to each of those actions,” he told the group.

AI governance is absolutely a priority for companies, but it takes a much more muted tone than the heated public debate involving research pauses and slowdowns and fear of agents slipping control of humans and unlocking massive security risks up to and including a mass extinction event.

Sophia Velastegui, CEO of Velastegui Ventures, said company boards are under pressure to maintain accountability and governance while meeting the market’s demand to go faster. A former chief AI officer at Microsoft’s business application group, she said that deploying AI without updating older systems is problematic, given that AI amplifies strengths and weaknesses in a system. 

She advised understanding AI agents as “personas” with their own strengths and weaknesses as a principle of management. 

During a panel discussion later in the event, the WSJ Leadership Institute’s Ben Ashwell asked Simile co-founder and CEO Joon Sung Park if we’re all going to die because of AI.

“No, I don’t think so,” said Park, whose AI company builds human-grounded simulations. “But I also think that the opportunity we have, the risk we have, is real. It’s a good time for this debate, but I think we will be OK.”

Do you think that AI is going to kill us all in the next decade? Send your feedback to me at steven.rosenbush@wsj.com (if you’re reading this in your inbox, you can just hit reply).

Who’s driving the AI strategy at FedEx? According to CEO Raj Subramaniam, he is. And Chief Digital and Information Officer Vishal Talwar said he’s “100% cool” with that. 

“AI is the defining technology of the era, so the CEO has to drive it,” Subramaniam said during a Monday night panel. “The gap between business and technology is fast going away. You can’t have a business strategy because technology and the reverse is also the case.” 

When Subramaniam hired Talwar a year ago he said he was looking for someone who understood his vision: leveraging the two petabytes of data FedEx generates daily to invest in AI that not only drives efficiencies, but also differentiated experiences for customers and ultimately new business models for FedEx. But also important was someone who could actually execute on that vision and understood the culture of the company.

Talwar, who was coming off a decade at Accenture, said he was excited to work with a CEO with such technology focused ambitions.

“It’s rare that you get, as a CIO, an opportunity where the CEO firmly, organically believes that technology is a feature and not a bug…and is looking to blur the lines, if not completely erase the lines, between business and technology.” 

Do they ever disagree? All the time, Talwar said. “We have so many debates,” he added. And it isn’t always obvious who’s right. “Sometimes he [wins the debate], sometimes I do. But, I would say, FedEx always wins.” 

Still, one area where they’re not disagreeing—even amid growing questions about the threat AI poses to humanity—is their belief that the technology can be a force for good, both at FedEx and society at large: “I think this one I can say safely for both of us, there is no debate,” Talwar said. “We both think that it’s in our hands, and it’s up to us to apply it for good.”

More takeaways from Monday’s sessions:

Setting AI budgets has turned out to be—and continues to be—a confusing nightmare.
“I’d be very surprised if people can say that they budgeted accurately in 2026. I don’t think most people did.”—Ruchir Swarup, partner and CIO at KKR

Seasoned execs have a surprising amount to learn about AI from the younger generation.
“My teachers now are my younger professionals that are actually closer to the technology, that are willing to take more chances with the technology. And they’re teaching more experienced people, including myself.”—Neil Dhar, senior vice president, IBM Consulting

Corporate boards are more invested than ever in shaping AI strategy—and it’s showing up in executive searches.
“I’ve never seen so much board involvement in a search process as I have for the last two years. I’m working on a Fortune 100 search right now where five board members are involved. I’ve never seen that before.”—Katie Shannon, global practice managing partner, AI & Technology Officer Practice at Heidrick & Struggles

Inventing new C-suite titles focused on delivering specific technologies might be a really bad idea (sorry Chief AI Officers!). “How many of you lived through the “Chief Cloud Officer” time? Dumbest thing in the world, right? The problem with associating a person with a technology is that that technology suddenly becomes a solution for everything, because their bonus depends on the adoption of that technology…I think we should have chief roles for outcomes.”—Tilak Mandadi, executive vice president, ventures and chief experience and technology officer at CVS

• Amid Monday’s AI scares, shares of Microsoft, Alphabet, and Meta Platforms all rose 2% or more—a sign, says the WSJ, that the market still expects these companies to remain among the world’s most profitable even if AI investment were to slow. Barron’s adds that the three firms may actually stand to gain from a slowdown, since they’re currently trailing OpenAI and Anthropic in both technology and model revenue.

• Microsoft has published a provisional code of conduct governing how it will train future AI models, joining a growing list of companies weighing in on limits for advanced AI systems, the WSJ reports. Under the code, Microsoft’s AI models must refuse requests tied to weapons development, avoid generating violent or sexually explicit content, and decline to assist with acquiring dangerous substances, among other limits.