00:18:48 · Kevin B. Boyd, Grace Gong
AI can be incredibly helpful at looking through, you know, many of the photographs of old stone tablets and cartouches on a wall and say, okay, I think there's actually a relationship between this ancient language and these newer languages and finding some of that. That's really interesting and helps us understand the evolution of languages that are the precursors to languages that we have today. Speaking [00:19:21] S0: of, I feel like living, especially like I feel like building on top of campus, like I feel like there is a lot of question about like, you know, identity and cybersecurity. I feel like those are like some really huge concerns for, you know, like whenever people say, oh, like, like somewhere that like the data league, a lot of times is universities since, like, it is... Where, like, a lot of, like, the student datas are, like, sitting on top of. What what are things that you would do to decide on, like, you know, which kind of like, whether it's like cybersecurity product that you guys are like interested in or, you know, when we went to that you like the event or like the private thing in that pile, like, it's very, I would say, like a lot of it it's like, you know, on edge device and, you know, like, I guess like telecom related things. And I wonder, like, what are, how does the decision making process look like for you guys to think about like from the security perspective?
00:20:29 · Kevin B. Boyd
[00:20:29] S1: So security is incredibly important for us. And I think, you know, in an AI world, it's security, it's privacy, it's data governance, and all of those things together. When we look at security tools, you know, my CISO and I often talk about the fact that it's it's such a target rich environment. There are so many opportunities. Every single week, there are a dozen people who are trying to sell us a new security tool, a new security solution. It's really easy to get sort of caught up in what's the latest cool cybersecurity tool. And I think what, you know, the way we think about it is we've gotta step back and really focus on what are the requirements. What's what's the problem that we're trying to solve? And then from there, we usually take a pretty structured process. So we'll look at, you know, what are the tools that are available in the market in that particular space. Yeah. We'll compare the the features, the functions. We'll look at the stability of the company, all of that. But it is a space that's moving rapidly, and I think you know AI is really changing the game for cybersecurity. And it's both a positive and a negative. You know, the positives are that our security team has much better tools that allow us to analyze vulnerabilities, find problems faster. But the flip side in the security space is that the bad guys, the cyber attackers, also have AI and that's allowing them to move much faster than what we'd ever seen before.
00:22:14 · Grace Gong, Kevin B. Boyd
[00:22:14] S0: When I guess like when you were thinking about like, know, since like you hear about these pitches every day, is there like a particular product that like everybody's selling you to? Or what kind of attack like you feel like are things that's like the most frequently happening and like that... That's like kind of like on top of your mind when you are evaluating the different things that they're trying to pitch you? [00:22:44] S1: I'm not sure I understand the question. So are you asking what are we being pitched most often? Yeah. You know, don't know that I have a good answer for that. I think it runs the gamut. For whatever reason, it seems like in the last few weeks, it's been a lot of tools that supposedly help with AI governance. And AI, what is essentially data loss prevention. So making sure that data isn't going into places or going out of places that it shouldn't be. [00:23:27] S0: Speaking of that, so like if you're building like a AI assistant that searches like the university documents, how do you make sure it's only like retrieve the information that's particularly, you know, that the user is like allowed to see?
00:23:27 · Grace Gong, Kevin B. Boyd
[00:23:27] S0: Speaking of that, so like if you're building like a AI assistant that searches like the university documents, how do you make sure it's only like retrieve the information that's particularly, you know, that the user is like allowed to see? [00:23:41] S1: Yeah, that's a great question. We've been having lots of conversations about that. And, you know, I think... I guess I would give you a couple different answers. So one of the things that is possible with Claude and with other tools like that is through the use of connectors, we can connect to, you know, various other... Or data storage, whether that's Box or OneDrive or Google Drive and things like that. I think generally those connectors are built in a... In such a [00:24:19] S0: way
00:24:19 · Grace Gong, Kevin B. Boyd
[00:24:19] S0: way [00:24:19] S1: that they're pretty good about people only being able to see what is available to them as a user. But I think one of the problems that exists is that often people have been given access to more than maybe what they actually need. And that hasn't been all that apparent in the past when they actually had to search through and scan through and look at things in some of these file storage systems. It becomes harder when all of a sudden the AI can see everything. And you know, if you ask it a question, it's going to pull back an answer based on everything that is available to it. And that's something where with many of these tools, we're trying to make sure that there isn't data oversharing, and that in tools where you have sensitive information, we're being intentional about maybe not making that sensitive information available to the AI. There's there's different ways to do that. So I think that's that's more on the... Some of these file sharing tools that are incredibly useful with AI. Then on the other side, when we're building agents, whether that's in finance or HR or others, so much of that is making sure that we're designing it, testing it, and really limiting access and putting appropriate guardrails in place so that we know exactly what it does have access to, what it doesn't have access to, and and we're and getting some warning if it seems to be exceeding guardrails in some [00:26:02] S0: way. When an AI is answering something wrong, so how do you figure out whether the problem was the motto or the retrieve document or the prompt?
00:26:02 · Grace Gong, Kevin B. Boyd
[00:26:02] S0: way. When an AI is answering something wrong, so how do you figure out whether the problem was the motto or the retrieve document or the prompt? [00:26:16] S1: So you're talking about less agentic in that question and more someone is interacting interactively with the tool? Is that what you mean? [00:26:28] S0: I would say, yeah, like more of like interacting since, you know, I feel like if you are building out, let's say like in the, yeah, I guess like also in the school context or like in whether it's like, you know, yeah, like, maybe, like, a faculty or, like, people are trying to pull diff... Different level of, like, information from the system.