What they’re saying: IT CEOs on agentic AI

IT executives believe that agentic AI may help software engineers and developers build products with more agility. That was a conclusion from Nvidia’s recent GTC conference in Washington, DC, where a roundtable of Big Tech CEOs discussed not only the rise of agentic AI, but also the need to secure tech stacks against next-generation threats.

Alongside Cognition’s CEO and co-founder Scott Wu and CrowdStrike’s CEO George Kurtz, the roundtable included Aravind Srinivas, the CEO, co-founder, and president of Perplexity, and the CEO of Abridge, Shiv Rao.

Step on the gas? Wu said that software engineers are simply faster if they’re “working with the best AI tools and doing the most with that.” In some cases, he suggested, one hour of an engineer’s time using the best tools “corresponds to about six to 10 hours without using the tools.”

Every engineering team has lots of projects to work on, Wu added, “but they have to choose four because that’s how things are with engineering…The ability to speed up and do a lot more is really exciting.”

When asked whether future generations of developer tools could make developers obsolete,

Wu believes it will always be up to humans to decide what computers should do. “Every software engineer, what we all love doing as programmers is going on and automating processes, and figuring out how to make this part faster, and make this clear and simpler, and so on,” he said.

Step on the brake. With the rise of agentic AI, Kurtz said that cybersecurity has to “parallel the slope” of the technology innovation curve: “In every inflection point…you have to have security.”

According to Kurtz, one big challenge is AI enabling more sophisticated threats.

From Kurtz’s perspective, data is the key to solving “almost every security use case,” meaning the more data an organization has, the more problems it can resolve. If that’s the case, Kurtz said, the only way for organizations to keep up is to supplement human security analysts with AI agents who can keep up with threats, provided they have enough data.

“AI seems to be a good opportunity to deal with lots of that data,” Kurtz said. “What we try to do and what we see in the adversary universe is the time has dramatically been cut for the adversary to actually find vulnerabilities, exploit them, get in and pivot.”

Rao added that, as Abridge’s operations are impacting a significant number of patients

But agentic AI isn’t a magic bullet for security, Kurtz said, adding that there isn’t one company or technology able to secure everything: “You’ve got to apply the right security technologies to each of those technologies, and then you gotta connect the dots across them.”