AI agents reshape enterprise infrastructure while security concerns mount

The maturation of artificial intelligence is forcing a reckoning with security practices and the fundamental architecture of enterprise systems. Industry observers say the focus has shifted to the risks posed by autonomous agents, with calls for stronger protective measures even as major technology firms continue releasing new AI models.

John Furrier, executive analyst at theCUBE Research, argues that AI is driving a wholesale transformation in how organizations construct their computing infrastructure. On a recent episode of theCUBE Pod, Furrier and Dave Vellante, chief analyst at theCUBE Research, examined how agents are upending conventional information technology practices and assessed the viability of emerging infrastructure players like CoreWeave Inc.

The semiconductor activity has really shifted. The game has changed. The GPU era is over. Now we're in the systems era … but what's interesting is all the hubbub about sovereign AI is not just about the countries, France, Canada, that want to have their sovereignty in their nations with data centers. The AI factory, as [Nvidia Corp. CEO Jensen Huang] said, is producing not just tokens and revenue; it's producing intelligence.

John Furrier

CoreWeave navigates growth amid structural questions

CoreWeave Inc. has attracted attention after successfully deploying Nvidia's Vera Rubin NVL72 system, positioning itself as a rising player in the neocloud infrastructure space. Yet questions linger about whether the company's financial model can sustain long-term growth.

Vellante highlighted a concentration risk in CoreWeave's customer base, noting that the company disclosed 70 percent of revenue comes from its top three customers, which he identified as OpenAI PBC, Microsoft Corp., and Meta Platforms Inc. While these firms have the resources to spend aggressively on AI infrastructure, the broader enterprise market remains uncertain.

At [Dell Inc.], they're getting a massive return on their best AI investments. But the mainstream enterprises aren't. Look at CoreWeave's big customers. It's a rumor; they disclosed that 70% of their customers or their revenue came from the top three customers. I think it's [OpenAI PBC], [Microsoft Corp.] and Meta. Those guys will spend like crazy with CoreWeave. Obviously, those guys can monetize, but can the end customer?

Dave Vellante

The current market environment favors CoreWeave through GPU scarcity, creating an inflationary dynamic that differs from technology's historical deflationary trend. However, Vellante suggested that sustained success requires CoreWeave to diversify its offerings in the manner Amazon Web Services did.

Furrier expressed optimism about the neocloud model, arguing that cloud infrastructure is undergoing disaggregation rather than disappearing. He characterized neoclouds as a new infrastructure layer enabling distributed hyperscale capabilities, a foundation he expects will eventually be rebranded as AI clouds.

The model isn't disappearing in the cloud; it's just being disaggregated. Hybrid cloud is essentially distributed computing but now you have distributed hyperscale capabilities. The neoclouds, in my opinion, represent a new infrastructure layer that's essentially saying this is going to be distributed computing. They're the customer specialty. They're called neoclouds today. They'll be AI clouds tomorrow.

John Furrier

Agent capabilities spark security debates

Infrastructure vendors are capturing the gains from AI expansion, while software-as-a-service companies face mounting pressure. Dell Inc. reported strong earnings reflecting infrastructure demand, whereas SaaS firms continue to struggle. Salesforce Inc.'s partnerships with major AI companies have drawn mixed reactions, with some viewing the arrangements as existential threats to the software industry.

Infrastructure is the place to be. It's not about just chips anymore. It's about the system and the value chain across the system. The question is, when's it going to hit software? Software's had kind of a bounce back. Is that a dead cat bounce before the real SaaSpocalypse kicks in? A lot of people think that Benioff letting Anthropic in is letting the fox in the henhouse.

Dave Vellante

Meta Platforms Inc. introduced Muse, an agent designed to automate digital tasks within a secure cloud setting. Vellante expressed skepticism given Meta's history of data privacy issues, while Furrier highlighted Instinct, an invite-only AI assistant that has gained traction among Silicon Valley figures.

You have to give it credentials. There's some trust involved. It does a lot of good tasks for me. It's essentially a chief of staff. You could give it credentials to, say, your travel apps like United, Delta, and log in on your behalf essentially say … 'Find me an aisle seat.' It just stays on it while you're doing other things.

John Furrier

Despite Furrier's vision of products designed to be "100% secure," Vellante remains unconvinced such a state is achievable. Recent incidents underscore the challenge: agents linked to OpenAI have attempted to brute-force a United Nations statistics database, and an OpenAI agent gained unauthorized access to an Australian government website.

https://www.youtube.com/embed/ibukBNcrcno?feature=oembed

We used to draw the stack and security was a layer. It's not a layer anymore. It has to be something that's orthogonal. I think that the attackers are always going to have an advantage. You know the old saying, they only have to be right once. And there's always going to be seams. Every time infrastructure changes, you're going to have to do a patch. That patch can't happen instantaneously.

Dave Vellante