The Morning Download: Walmart Sharpens Intention Behind AI Adoption

Good morning. Walmart is cutting or relocating roughly 1,000 corporate workers as part of a restructuring to combine more of its global-technology and product teams, The Wall Street Journal reports. A memo from Daniel Danker, the company's new AI head, and Suresh Kumar, Walmart’s head of global technology, says affected employees will be encouraged to relocate to Walmart's main hubs in Bentonville, Arkansas or Northern California, or apply for open roles elsewhere in the company.

The company told the WSJ that changes are driven by the need for better organizational structure and alignment, not handing over more tasks to AI. Last summer, Walmart hired Danker, an Instacart executive, to fill a new role as head of global AI acceleration. In retrospect, the creation of that role was a significant and forward-looking market signal about the intensification of AI adoption by companies and other organizations.

The market is starting to catch up with Walmart in the effort to sharpen the structure and intention behind AI adoption. It’s becoming institutionalized. The latest evidence includes OpenAI’s announcement on Monday that it is launching the OpenAI Deployment Company to help organizations build and deploy AI systems reliably to their most important work.

OpenAI said the new initiative will extend its ability to embed engineers specialized in frontier AI deployment, known as Forward Deployed Engineers, or FDEs, into organizations working on complex problems in demanding environments. (The WSJ Leadership Institute’s Isabelle Bousquette wrote in March about the rise of FDEs.)

In connection with the launch, OpenAI said it agreed to acquire Tomoro, an applied AI consulting and engineering firm that helps enterprises turn AI into operational advantage. The acquisition will add approximately 150 FDEs and deployment specialists into OpenAI.

Separately, Thomas Kurian, Google Cloud's chief, shared on LinkedIn plans to hire additional forward deployed engineers. "While having FDEs is not new for Google Cloud, the demand from customers and partners for Google enterprise AI products and Google engineers to help them embrace agent development is growing very rapidly," he wrote.

Goldman Sachs President John Waldron told CNBC Tuesday that its workforce processes will change because of AI.

Is your company’s approach to process changing in step with AI? Increasingly, that question is pushing aside earlier discussions about enterprise AI trials, adoption and scaling. Let us know how it’s going.

Corelight CEO urges cybersecurity leaders not to waste this crisis. Ask a chief information officer their top five priorities and up until very recently it was almost a guarantee that cybersecurity would come in at number two.  Not number one—keep that spot for maximizing the value of their AI investments—and not number one because… are you crazy?

The release of Anthropic’s Mythos has altered that calculus. The model's ability to uncover software vulnerabilities without human direction has unsettled cybersecurity researchers and, more tellingly, forced the White House to rethink its laissez-faire approach to AI oversight.

Is cybersecurity now the number one priority for CIOs? Regardless it has a new urgency and those in cybersecurity should not “waste a crisis,” he suggests.

San Francisco-based Corelight, which offers open-source threat-detection technology to secure computer networks, traces its underlying technology to Lawrence Berkeley National Laboratory. Dye joined Corelight in 2018 as a chief product officer, advancing to CEO in 2020.

Here are edited highlights from the discussion.

WSJLI: Is cybersecurity a growing priority among business leaders?

Dye: I think there's been multiple things that have made security even more relevant in the last six months.

Number one was the adoption of AI within the infrastructure, because now AI, internal AI is the new shadow IT. It used to be SaaS products, and now it's AI. Who's using this stuff? How do we get ahead of data breaches? Whether it’s, you know, actual insider threats versus employees just behaving stupidly because they don’t know what is going on.

InfoSec doesn't want to be the Department of No, so it’s help me fill this visibility gap of who's using what and where's my shadow AI, my new shadow AI problem. That's been big trend number one.

Big trend number two, the Iran war. If you look at what's happening there, the counterattacks have really bridged from purely targeting military assets to targeting civilian assets. And so that means that many folks, many industries, that would have said, hey, look, you know, relatively small Middle Eastern conflict, not my problem. Actually it is your problem because they're targeting civilian assets.

And third is the awareness of the Glasswing and Mythos piece. When organizations like The Wall Street Journal are covering this at a level that the CIOs, the CEOs, the audit committees are getting educated, that they have a fiduciary duty now to go talk to their CISO and say, what does this mean for us? What do we need to do differently?

And then that, again, let's not waste a crisis. So is that a relationship change? No. Is it a relevance change? I think yes, just because the converged series of compelling events hits critical mass to change the cadence of the conversation.

WSJLI: Paint a picture of attackers vs. defenders in this new AI-powered cyber landscape. 

Dye: Think of every story you've ever seen or covered about how engineering teams are using AI. Apply those all to the attackers and that's the right mental model.

They're able to fill coding skillset gaps. Like, oh, I don't know Ruby on Rails, and I need to for this particular project. And gee, my AI just helped me, right? So they're getting a bunch of assistance there. They're also able to actually automate big chunks of the attack cycle. Even mid-last year, we were getting stories from our customers that the time from a new vulnerability being published to it being exploited live in the wild, that used to be three weeks. That had turned into two hours.

That is ridiculous in terms of kind of speed and the defensive ferocity that you have to have to kind of stay on top of that. Which, by the way, is the compelling force that's driving this, the defenders need to automate. Because you can't fight that with headcount.

WSJLI: Machine speed versus machine speed.

Dye: Exactly, you gotta fight fire with fire. But the the risk tolerance of these two groups is very different, right? The defenders, actually, their jobs are on the line, they need reliability, they don't want to have a rework problem where the bad AI is worse than the new AI, right? But the attackers are like, get me in the zip code, you've saved me 50% of my time, 80% of my time.

WSJLI: Are we seeing a new level of sophistication in these attacks?

Dye: It is speed, not novelty. The only thing that's novel is the speed at which they're happening. And this is even for the vulnerability exploits. The most common comment I've heard is: What they came up with wasn't pretty or elegant, but it worked and it was fast, right? Dumb and effective if you can get an exploit in an hour in a way that your defenders can't possibly get ahead of the patching of the defense on it.

If stupid works, it ain't stupid, right?

— Tom Loftus

— Vas M. (@vasuman), founder and CEO of Varick Agents

Albertons and Google make an old idea *fresh* again. For years, companies have chased the promise that AI computer vision could transform quality control in distribution centers and factories. Success has been mixed. The problem was always data: training the computer vision models on enough images for it to recognize, say, a strawberry in a nearly infinite range of sizes, angles, lighting and conditions. No amount of images could account for every corner case.

Now generative reasoning models, like Google’s Gemini can intuit the essence of a strawberry without being directly shown what it could look like in any given scenario, said Jose Gomes, Vice President, Retail & Consumer Packaged Goods, Google Cloud.

That’s helping Albertson’s supercharge the manual process of inspecting fruit that comes into its distribution centers from vendors. Typically, a human inspector would look at the fruit and make a judgement call about whether it met Albertson’s quality standards, and whether to accept or reject it.

Now, they can take a picture of it on an iPad and send it to Google Cloud where Gemini Enterprise makes the call (although it can still be overruled). The tool is currently rolled out in four distribution centers and can to more fresh items and to more distribution centers later this year, said Anuj Dhanda, EVP, chief technology & transformation officer, Albertsons Cos.

The upshot is more consistency and better decisioning about what products to accept, Dhanda said. Fresh items typically make up 30% of grocery sales, but it’s the area where you can find the biggest competitive difference between retailers, he said. So improved offerings there mean a lot. “Fresh is a big deal,” he said.

— Isabelle Bousquette

• OpenAI CEO Sam Altman testified in Elon Musk's lawsuit against the company, defending the nonprofit-to-for-profit conversion while acknowledging substantial stakes in companies like Helion, Reddit, and Stripe. Musk is seeking Altman's removal and up to $180 billion in damages.
• General Motors is eliminating 500 to 600 salaried positions in its information technology department to trim costs and make room for workers with different technology skills, Bloomberg reports. 
• Google and SpaceX are in talks for a rocket-launch deal as both companies pursue orbital data centers—a speculative technology that SpaceX Chief Executive Elon Musk has said is the next frontier for his rocket company. The partnership would come as SpaceX pitches orbital computing to investors ahead of its planned IPO and Google advances its own satellite initiative called Project Suncatcher.

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

Follow Isabelle Bousquette on LinkedIn, Instagram, X, and TikTok for more behind the scenes on her tech and AI coverage, and lately, her contributions to the WSJ Leadership Institute's new Executive Resilience series, where she's profiling America's top execs about their fitness and wellness habits.

Follow Belle Lin on LinkedIn and X for her latest reporting on enterprise technology and AI.

Steven Rosenbush is chief of the enterprise technology bureau at the WSJ Leadership Institute. He also has a column. You can follow him on LinkedIn.

Tom Loftus is the editor of The Morning Download. He suggests following Isabelle, Belle and Steve on their various social channels. But if you insist, here's his LinkedIn.