The Morning Download: Corporate AI Adoption Deepens

Good morning. The WSJ Technology Council Summit in Silicon Valley illuminated where business adoption of AI stands today. And that matters, because corporate AI budgets are central to the success of model developers, chip companies and data-center builders in the global AI spending and investment boom.

Most of the public’s understanding of corporate AI adoption comes from surveys and studies. They can be useful, but they’re abstractions based on what has happened in the past or what people say they will do in the future.

On Tuesday and Wednesday, we had the opportunity to speak directly with CIOs and other business technology leaders about where AI adoption stands in their companies right now. And there’s no more second-guessing whether the bots are here to stay, Belle Lin and I write in our report from the event.

The leading edge of AI adoption includes a large cohort of big enterprises. The early adopters often come from regulated industries such as financial services and life sciences, which have lots of data, plenty of cash to fund investments and a long-established capacity to manage risk, of which AI is one of many species.

“You can't do anything hardly anymore without your employees using AI,” Kathy Kay, CIO of global financial services firm Principal Financial Group said Wednesday during an interview on the sidelines of the summit.

Of course, adoption is still in the early stages and can face a range of obstacles. While AI agents that can perform tasks of increasing complexity are getting deployed inside companies, many challenges still remain—from concerns about cybersecurity to governance and employees who fear that AI will take their jobs.

Stephen Carvelli, chief technology officer of Sonic Automotive, said the publicly traded dealership chain has had “pretty good success” with AI agents that improve the customer experience. You can, with agents, be there 24/7, Carvelli said during a panel discussion that also included Bank of New York Mellon Chief Information Officer Leigh-Ann Russell.

The bank has 130 AI-powered “digital employees” which have human managers, according to Russell.

Taken together, those businesses show the reach of AI agents has extended far beyond Silicon Valley. Principal Financial Group is based in Des Moines and dates back to 1879. Sonic Automotive is based in Charlotte, N.C. and was incorporated in 1997. BNY is based in New York and was founded in 1784.

Silicon Valley is the epicenter of AI, though. Alex Balazs, Intuit’s chief technology officer, said on the sidelines of Wednesday’s event that the Mountain View, Calif.-based company has found success with AI agents that can close books for the software company. It also uses customer-facing agents and has a strategic partnership with OpenAI.

Balazs told Belle and I that “the days of AI assisting you with coding have become you are assisting the AI with coding.”

It’s important not to get carried away in the fervor, though. Wall Street’s top brass have sought to soothe investor jitters over AI’s threat to software, but many attendees at the summit said they are more concerned about putting AI agents to work inside their companies than replacing their existing software with AI.

“There are issues with allowing agents to be able to do their work properly, even authentication,” Carvelli said on a panel. “The tools need to be a certain level of security, a certain level of capability.”

If we didn’t see you at the summit, we hope you will join us next time. In the meantime, have your AI personal assistant let our AI personal assistant know what, statistically speaking, you probably think.

Leadership lessons. Intuit Chief Executive Sasan Goodarzi has watched his company’s market value plunge in recent weeks as an AI-driven selloff rattled traditional software stocks. He’s not fazed. The Iranian-born executive, who moved to the U.S. at 9, endured bullying and flunked out before landing in community college, says he’s faced tougher tests.

“That is really what shaped me," he said of his childhood. "It’s the notion of building grit and resilience that tomorrow will be a better day.” 

From CTO to CEO to founder/CEO. Before Parag Agrawal launched Parallel Web Systems—an infrastructure play for an agent-driven internet—he served as CTO and later CEO of Twitter. The arc, he says, reshaped his approach to leadership.

Being CTO at a “post-product-market-fit company” meant “adapting yourself to that circumstance.” Becoming CEO required “adapting the company to you and who you are,” he said Wednesday at the WSJ Technology Council Summit. And becoming a founder? It’s “finding your inner calling and really embracing, taking it to the next level,” he said.

🎥 Former Twitter CEO Reflects on Platform's Legacy. Former Twitter CEO and Parallel Web Systems CEO Parag Agrawal reflects on the platform's legacy, saying it's "special to have built something that is resilient beyond yourself." 

It's not just about LLMs or AI chips or even AI agents. According to Leigh-Ann Russell, CIO and global head of engineering at BNY, what really matters is “coherence.” It’s her word of the year. 

"The great thing about AI and the threat with AI is you can just no longer throw money at a problem," she said. "You can't get more compute power, more LLMs, and everything's going to magically appear," she said. "You have to have coherence in your organization." 

That means data must align. Permissions must align. And most critically, "the business outcomes have to line up with the inputs from the engineering organization."

And you better get it quick.  “I call AI years 'dog years' now because how fast the pace is moving," she said.

🎥 The New Speed of Business: Why AI Adoption Is Moving in ‘Dog Years.’ BNY CIO Leigh-Ann Russell and Sonic Automotive CTO Stephen Carvelli discuss the rapid pace of artificial intelligence innovation and the move toward customer-facing AI agents. 

Rethinking build vs. buy. It’s not either-or, but what’s core and what’s not, said Winston Weinberg, CEO of legal AI provider Harvey AI. “There was much more we wanted to self build in 2024 and less in 2025 and now it is significantly less,” he said. "Most enterprises I know are puting all of their build resoruces iinto the core competancesis of what their enterprise is."

Sequoia Capital Partner's Pat Grady offered a tactical tip: Track token usage to spot where third-party tools make sense. "If you pay very close attention to the people who are consuming the most tokens, they're going to lead you to what applications you should probably think about buying."

🎥 Why AI Is In Its ‘February 2020’ Moment for Business. Sequoia Capital Partner Pat Grady analogizes the AI boom to the onset of Covid-19, warning that the technology will be "much bigger" and more durable, requiring leaders to shift from playing defense to playing offense. 

Beware the 'Turing Trap.' The classic Turing Test asks whether AI can perfectly imitate humans. Ignore it, said Erik Brynjolfsson, co-founder of Workhelix and a professor at the Stanford Institute for Human-Centered AI. "It's too low a ceiling," he said, adding that mimicry isn’t the goal, amplification is. Framing AI as a replacement tool is a “sociological nightmare,” he warned, because employees resist systems designed to supplant them.

The smarter path? “Centaur benchmarks” that measure human-plus-AI performance. "I think that's the path towards, you know, having a higher ceiling, having more people embrace it," he said.

🎥 Erik Brynjolfsson: Productivity is 'Much Higher' Due to AI
The Stanford economist argues that the "J Curve" of AI adoption is turning upward. Brynjolfsson estimates 2025 productivity growth at 2.7%—double the previous decade’s average. 

From SaaS to outcomes. The real shift isn’t just better AI, but a new business model, according to Bret Taylor, co-founder and CEO of Sierra.  As agents autonomously complete workflows, value moves from systems of record to systems that produce results. "If you look at the evolution of technology, it's gotten more and more accountable over time from perpetual licensed on premises software to a subscription to cloud software, to now outcomes based pricing for agents," he said.

🎥 Why Software Pricing May Move to Pay-Per-Outcome
Sierra Co-founder and CEO Bret Taylor discusses the future of software billing, predicting a move away from monthly licenses to a model where companies only pay when an AI agent successfully completes a job or closes a sale. 

Wednesday brought another installment of the AI jitters. Shares of Bank of America, JPMorgan and Citigroup each dropped more than 2%, while brokerage stocks like Charles Schwab and Robinhood continued their slide, the Journal reports.

Software names Salesforce and Intuit each ended the day down more than 4%.

The selloff followed news that fintech firm Altruist launched an AI tool to generate personalized tax strategies without manual entry, fueling fears of disruption in brokerage and wealth management. The rush by investors echoed a similar retreat in software stocks last week after investors homed in on AI news involving Anthropic.

A message to love, not fear the AI. AppLovin CEO Adam Foroughi dismissed fears that AI threatens the company’s ad-tech platform, arguing stronger AI models will fuel, not hurt, growth. The company, which provides software and AI for mobile app developers, guided for sequential growth, driven by gaming and expanding e-commerce clients. Through market close, shares have slid 32% this year.

Anthropic enters midterm-election showdown. The AI startup is donating $20 million to Public First, a group pushing for guardrails and transparency from AI developers, the WSJ reports Public First was set up to oppose a super political-action committee network launched by executives at OpenAI and venture firm Andreessen Horowitz to fight what they say is burdensome regulation.

Musk announces xAI reorganization, staff departures AI on the moon. Elon Musk announced a reorganization of xAI, just days after merging it with SpaceX, the Journal reports. During a company all-hands meeting Wednesday, Musk confirmed that the reorganization would involve the departure of some employees. Two of xAI’s co-founders confirmed in X posts that they were among those leaving.

Musk also unveiled a new organizational structure for xAI that divides the company into four main teams: Grok, its AI chatbot; another devoted to a coding-specific AI model; a third working on its image-generation model; and a fourth pursuing a project dubbed Macrohard, aimed at simulating software products with AI.

A new focus on Mars moon. On Tuesday, Musk told employees that the company needs a lunar factory and a catapult to launch AI satellites, expanding on his plan to merge xAI with SpaceX and create AI data centers in space. He framed the moon as a steppingstone to Mars and eventual interstellar exploration, according to the New York Times.

Over the past year, demand for healthcare workers has quietly propped up the labor market as other sectors reined in hiring or even shed jobs. On Wednesday, the full power of healthcare’s role burst into full view, and marked a clear shift in a labor market now geared toward the hard, often physical, work of caring for America’s aging population. (WSJ)

The GOP-led House passed a resolution Wednesday designed to roll back President Trump’s tariffs on Canada, as a half-dozen Republicans joined Democrats in rebuking the administration’s signature economic policy. (WSJ)

The Chinese owners of the famed Waldorf Astoria in New York City are preparing to put the hotel up for sale, only months after it reopened following a multibillion-dollar, much-delayed overhaul. (WSJ)

McDonald’s said that its multiyear quest to make its food more affordable is working. The world’s largest burger chain reported that global same-store sales rose 5.7% in the three months ended Dec. 31, outpacing analysts’ expectations for the quarter. (WSJ)

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

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