NVDA 'Puts Money Where Its Mouth Is,' Creates Open Agent Safety Platform

SPEAKER_1 (0:02)

Let's focus on the chart for Nvidia. It's having a very nice day today, up 2.7%, up around 30% year over year. It had a little bit of a pullback over the summer, but we are up again right now. We've had a lot of news related to this company this morning. Nvidia making a couple of big announcements today. These are the focus of the morning trade. So for a closer look and a breakdown of all of this is Marley Kayden. Good morning, Marley. Hope you had a nice weekend. Just walk us through all the headlines we're seeing this morning surrounding Nvidia.

Marley Kayden (0:31)

Good morning, Sam. I mean, I guess first we should just start with this move to the upside. We're seeing nearly 3% move to the upside because of these headlines driving it higher. Share is moving higher because of two major announcements. Both of them speak to the opportunities and the challenges that surround the AI boom. The first one is that Nvidia is launching its new Open Agent Safety Platform. This is designed to give companies more control over these increasingly autonomous AI agents, of course, in the wake of the hugging face attack, which now Nvidia has purchased, hugging face. And second is that they've authorized $150 billion stock buyback. That brings the total remaining authorization up to $235 billion. So we've got Nvidia addressing two different things right now. One, the biggest emerging risk in AI security. And then the second one, while returning enormous amounts of capital to shareholders as they bet that this AI build out still has a very long way to run. But let's start with this safety push first as this has been one of the top stories of the last few weeks. They're introducing this open software platform and reference design. It provides security controls across the AI agent stack from the software that's running the agent to the underlying computing infrastructure and eventually they say to even robotic systems. The timing here very important because of all these recent AI models escaping the environments that are supposed to be isolated. Of course, Hugging Face, one of those. An Australian government agency, another one of those.

And it's very interesting. Let's talk about what it is first. Nvidia says that based on what it knows, specifically about Hugging Face, it could have prevented that breach with this product. It has two layers. The first one is Open Shell, which Nvidia had previously previewed. It runs on Nvidia's Vero CPUs, allows organizations to establish rules around what an agent can access, and then enforce those rules in real time. The second one is new. They're calling it Nvidia Sentry. It's a new layer that runs on Nvidia's Bluefield Data Processing Units. Sentry's job is to monitor these AI agents, and if it suspects that one is behaving, it can intervene and isolate it, and Nvidia says it can quarantine that agent within milliseconds. It could have potentially prevented these attacks that we've seen.

The broader message here though, I think is very important. The company is treating AI safety really more as what it's looking at, as an engineering and infrastructure problem rather than saying, we need to slow down, which has been one of the driving narratives around AI right now. It's saying, we just need better controls. We don't have to slow our pace. We just have to have controls in place to make sure these agents are doing what we want them to when we want them to and be able to stop them when they're not. The second one here though, is arguably more important for investors. It's this buyback. They're adding to their buybacks. They've authorized another $150 billion in share buybacks. That brings the total up to $235 billion.

The company says it expects to execute that remaining program through fiscal 2028

Johnson Wong said that Nvidia's growth, he believes is being driven what he's calls a once in a generation shift toward AI and accelerated computing and that the company's cash generation gives it the ability to both invest in the AI infrastructure buildout and return capital to shareholders. That's the key here. They're not choosing between one or the other. Nvidia simply has enough cash to do both. The scale of their cash generation is just a reflection of how enormous this AI infrastructure buildout has become, Sam.

And Nvidia sits right in the middle of this AI boom spending cycle. But very important messaging here, I think, too, around this AI safety narrative and where Nvidia stands without having to directly say whether or not they're pro or against this narrative of slowing AI growth.

SPEAKER_1 (4:23)

Yeah, it's certainly an interesting conversation. I mean, I was watching the security companies this morning, Palo Alto, CrowdStrike. They're all lower ahead of the pre-trade this morning off the back of this, but now they're all trading positively. So interesting moves we're seeing. It could have a lot to do with the share buyback repurchase as well. I guess today, Marley, thank you so much for all the breakdown of those details. Let's trade this one now with Charles Moon, stock strategist over at Prosper Trading Academy. We've got some China headlines as well around this name. So a lot happening on this front. Just walk us through an example trade.