Good morning, CIOs. In the past, success in the services business was defined by speed and efficiency, but that isn't enough anymore, according to Liz Centoni, chief customer experience officer at Cisco.

"Customers are looking for relevance, and that's where they're looking at us,” she told me during a recent conversation at company headquarters in San Jose, Calif.

We spoke about how AI has both raised customer expectations and equipped the tech giant with the capacity to do more. She also spoke about how she addresses the challenges of driving AI adoption across her organization, which is close to one quarter of the tech company’s workforce. Read on for highlights of that conversation, and much more.

WSJ Leadership Institute: How are customer priorities changing? 

Centoni: Customers are looking for relevance, and that's where they're looking at us. We've got 40 years of their data. We've got massive amounts of human insights where we've solved many of these problems many times over... What customers are (saying) is you have all this data, you know my environment, I want you to bring that context, I want you to tell me what potentially is going to happen before it actually happens. I want you to alert me to risk in my network. That's what they expect from us given all we know about their environment. So I feel like automation, efficiency, speed, being fast—all great, but what they're looking for now is beyond that.

WSJLI: To what extent do AI agents play a role in that effort?

Centoni: Agents take on some of the tasks, not the full workload.  We're not there. We're not making the workload fully autonomous. We're making tasks within the workflow autonomous without needing a human to touch it. That's the difference for me between what's AI and what's agent work.

So, the data collection work -- I don't need to go in and do this.The synthesis, the analysis, I don't need to go do that. Finding the risk from five different places. If it's automated, I don't need to do that. 

We need to prompt the agent to do that. Our future is, it's so immersed in the workflow, where the sales agent is talking with the renewals agent for example and I don't need to prompt it.

WSJLI: Is there a return on this investment?

Centoni: In customer support, the percentage of complex cases resolved in a single contact has jumped from 28% to 40% under the new AI systems, with an ultimate goal of reaching 75%. As a direct result, the overall cost to serve customers has decreased, improving gross margins. On the software side, the AI agents have successfully reduced churn dollars and increased overall renewal rates.

WSJLI: You oversee 20,000 people. How have their roles changed over the last 18 months and how are they handling that?

Centoni: Adoption of the renewals agent currently sits between 35% and 40%. Employees will instinctively revert to comfortable, manual workflows if given the option, especially if an AI tool is only 90% accurate.

There's big change-management work that's needed, because you're dealing with people and I can say a thousand times that I'm not looking to let go of people, I'm looking to create capacity, it still generates fear.  I don't sugarcoat, I tell them this is a mandate, we are going to do it. But it's helping people redefine what their roles are. They need to upskill as well. I think that change management and people and culture and work transformation is a lot harder to do than the technology.

The biggest AI companies for a long time amassed marketing mileage and new customers by touting the nearly godlike transformational potential of their tools. The dangers they talked about felt like far-off science fiction.

Now, as some businesses, investors and individuals become more concerned about who will be swamped in that transformation, the companies are making a pivot to trying to help manage the downside, the WSJ reports. 

“We do feel an urgency to this conversation,” Chris Lehane, OpenAI’s chief global affairs officer, told the WSJ.

• OpenAI this week published a populist wish list of policy proposals that zero in on worries like job replacement and wealth concentration, floating such ideas as a four-day workweek and an AI-invested public-wealth fund distributed to citizens.
• Rival Anthropic has been signing partnerships and building tools for such sectors as consulting and software, where share prices have been whacked by investor worries that they will be replaced by AI. Anthropic’s efforts have helped push back up shares of tech companies including LegalZoom.com.
• Anthropic and OpenAI are each pursuing ventures to help private equity, a big owner of companies in sectors ripe for disruption, with AI transformation. 

Anthropic set to preview new AI model to ward off cyberthreats. Mythos has proven to be so capable at things like finding and exploiting software bugs that it has no plans to release it to the general public, the Journal reports.

Instead, the company is making a preview model available to about 50 companies and organizations that maintain critical infrastructure, including Amazon, Microsoft, Apple, Google and the Linux Foundation.

Buggin' out. AI's ability to find, exploit bugs has been ramping up. Late last year, Stanford University researchers found that AI software was almost as good as humans at finding and exploiting bugs on a real-world network. And earlier this year Claude Opus 4.6 found more high-severity bugs in the Firefox browser in two weeks than the rest of the world typically reports in two months.

Anthropic snags Microsoft exec to lead infrastructure. Eric Boyd previously oversaw Microsoft’s AI platform, leading a unit of 1,500 workers, Bloomberg reports, citing an organizational chart. 

Elon Musk files amendment to his OpenAI lawsuit. Musk is now asking that any damages he might win be awarded to OpenAI’s charitable arm rather than to himself. The amendment also calls for CEO Sam Altman to be removed as chairman of ithe nonprofit's board, WSJ reports. Musk is seeking more than $150 billion in damages from OpenAI and Microsoft arguing that OpenAI strayed from its nonprofit mission and defrauded him as a donor in seeking to convert it into a for-profit company. 

Intel partners with SpaceX, Tesla and xAI. The chipmaker said it will work with Musk's companies to “design, fabricate, and package ultra-high-performance chips at scale,” a win for Intel, which has struggled in recent years.

GoPro looks to slash headcount. The wearable camera maker said Tuesday its board approved a restructuring plan to slash costs, which will entail cutting 145 employees. GoPro’s head count at the end of the first quarter was 631 workers, it said.

President Trump said he agreed to suspend attacks on Iran for two weeks subject to the immediate reopening of the Strait of Hormuz. (WSJ)

A growing number of Democrats are calling for the 25th Amendment to be invoked or other means to remove Trump from office, following the president’s comments Tuesday morning that Iranian civilization “will die tonight” if no deal was reached by 8 p.m. ET. (WSJ)

President Trump has ousted Kristi Noem from the Department of Homeland Security, but his administration is hanging on to the controversial $70 million jet she leased during her tenure. (WSJ)

The WSJ Tech Council brings together CIOs, CTOs and CISOs advancing innovation and shaping the future. Join this trusted community where tech executives connect with peers to explore emerging trends and gain the perspective they need to stay ahead of disruption.

Request Information

The WSJ CIO Journal Team is Steven Rosenbush, Isabelle Bousquette, Angus Loten, and Belle Lin.

The editor, Tom Loftus, can be reached at thomas.loftus@wsj.com.