Learning Library

← Back to Library

Human vs AI Agent Identities

Key Points

  • The speaker introduces a discussion on AI agents and “agentic identities,” inviting open, non‑debative feedback from the audience on emerging industry questions.
  • Human employees are framed as physical beings belonging to organizational structures who follow a task lifecycle: receive → assess → plan steps → execute → learn and improve.
  • Traditional non‑human identities (NHIs) are described as purely digital, deterministic entities that perform tasks in a fixed, unchanging manner.
  • By contrasting human and non‑human identity models, the speaker sets the stage for exploring how AI agents might blend or redefine these characteristics in corporate settings.

Full Transcript

# Human vs AI Agent Identities **Source:** [https://www.youtube.com/watch?v=D9fp2OVeAIo](https://www.youtube.com/watch?v=D9fp2OVeAIo) **Duration:** 00:11:52 ## Summary - The speaker introduces a discussion on AI agents and “agentic identities,” inviting open, non‑debative feedback from the audience on emerging industry questions. - Human employees are framed as physical beings belonging to organizational structures who follow a task lifecycle: receive → assess → plan steps → execute → learn and improve. - Traditional non‑human identities (NHIs) are described as purely digital, deterministic entities that perform tasks in a fixed, unchanging manner. - By contrasting human and non‑human identity models, the speaker sets the stage for exploring how AI agents might blend or redefine these characteristics in corporate settings. ## Sections - [00:00:00](https://www.youtube.com/watch?v=D9fp2OVeAIo&t=0s) **Opening Discussion on AI Agent Identities** - The speaker invites an open, non‑debative dialogue about unresolved industry questions surrounding AI agents and their identities, framing it against the concrete, physical identities humans hold within corporations. - [00:03:03](https://www.youtube.com/watch?v=D9fp2OVeAIo&t=183s) **From Deterministic Bots to Adaptive Agents** - The speaker contrasts static, deterministic digital entities with AI agents that assess prompts, orchestrate task flows, and iteratively retrain based on performance metrics, making them behave more like humans. - [00:06:24](https://www.youtube.com/watch?v=D9fp2OVeAIo&t=384s) **Treating AI Agents as Coworkers** - The speaker interrogates whether organizational agents that perform support tasks should be regarded, managed, and listed in directories like human employees. - [00:09:31](https://www.youtube.com/watch?v=D9fp2OVeAIo&t=571s) **Cost and Governance of Scalable Agents** - The speaker highlights the CPU, network, and data costs of persistent agents, warns that scaling to thousands of agents will overwhelm current identity‑governance systems, and questions whether the existing IGA solution can handle the resulting approval and entitlement workload. ## Full Transcript
0:00Howdy, everyone. 0:02We're going to do something a little bit different today. 0:04We're still going to talk about AI and agents, 0:06and we've had a lot of conversations about that. 0:09But when we start thinking about AI and AI 0:12identities, there's actually some open questions that are still being debated out 0:16amongst companies and out in the industry. 0:19So I want to pose a set of questions today 0:22and engage everyone in feedback and comments 0:26so that we can actually have an open 0:28bit of a conversation about some of these questions and 0:31and invite you to give me your thoughts on how you think things are going. 0:36I don't want to make this a debate. 0:38I really am not looking for, 0:39you know, people to chime in with that you really are against something. 0:44I really am inviting everybody to just have an open conversation 0:49and think about some of the questions 0:51that are actually being posed out now 0:53about agents and agentic identities. 0:56All right. Before we get to that, 0:58I do want to revisit a little bit about identities 1:02and set that up as kind of a foundation for our questions and our conversation. 1:06So the first thing I want to talk about is humans 1:10and how we are represented 1:14as identities in corporations. 1:17Now, the first thing we have to understand, 1:20and this one is obvious, 1:23is that we have a physical existence. 1:25We we are here in a physical world, 1:29in organizations performing tasks and jobs, right? 1:33So we actually have a physical existence. Now, and 1:37I kind of just said this, the 1:38next thing that we have is we belong to organizations. 1:42So we belong to an enterprise or a company. 1:45We belong to a department, we belong to a business unit. 1:47But we do belong, in some way, 1:51to an organization within an enterprise. 1:55The next thing is 1:57I want to start thinking about how we perform tasks 2:02so as as employees. 2:04We perform tasks. 2:06And so the first thing that we typically do is we're given a task. 2:09And we have to assess 2:12what it is that I'm supposed to do. 2:15And once we assess it, we actually start breaking down the problem 2:19into a set of steps, 2:22especially for complex 2:24tasks like, okay, what is it that I need to do? 2:26What are the steps that I need to take? 2:28Once we've assessed the problem or the task, not a problem, and 2:32we break that down, the 2:33next thing we do is we execute. We 2:36actually perform the task. And 2:39then finally, we come back and we learn. All right. 2:41Did that work? How did I do at the task? Could 2:43I have done it better? 2:45Is there some way to change? And we learn from that. 2:47And we apply that the next time we do a task. 2:50All right, so that's humans. 2:52The next thing is our traditional non-humans. 2:59Or what we call 3:01NHIs, non-human identities. 3:03Now they're a little bit different. 3:05Obviously, they're digital. 3:09They have a digital existence. 3:11And traditionally, they also, when they perform a task, 3:17they're very deterministic. 3:21We know exactly what they're going to do and how they're going to do that. 3:25And it is typically unvarying, right? 3:29We really don't change 3:31how a non-human entity or identity works. 3:35It it does the task. 3:37So now let's think about agents. 3:38So we've been talking about agentic flows and agentic systems. 3:42I'm not going to talk about assistants. 3:44I really want to leave this to agents. 3:45When we think about AI and we think about agents. 3:49Now first obviously, they're not physical, right? 3:52They're digital, just like other non-human identities. 3:59But we really largely think they're going to belong as part of an organization. 4:03They're going to be in a business unit or a department performing task. 4:08And when they do perform a task, they're going to assess. 4:11This is really what we're trying to do with with AI and with agents. 4:15They look at what the prompt or the ask is. 4:18They assess what it is. 4:20They break it in down into tasks. How 4:21am I going to orchestrate this flow? How 4:23I'm going to perform the steps that I need. 4:25They execute on that flow, and then, they actually learn. 4:29We actually will retrain them. 4:31Well, look, what was the accuracy of the task they performed? 4:34Was it 70%? Was it 80%? Was it 90%? 4:37If it wasn't high enough, we'll go back and learn. So, 4:39in this case, they really start operating 4:43and behaving a lot like a human. 4:45And so when we talk about that, then 4:47so, as we start embracing agents into a workplace, 4:53there are questions that start arising around that. 4:56And so these are the questions that I want to pose 4:58to everyone to, kind of, chime in and give me your thoughts on this. 5:01The first one is: 5:04When we think about agents, is/are agents 5:11just software? 5:18Now obviously they are right. We say they're digital. 5:20We know that they're living in IT systems. 5:25And there are actually applications that we have running. 5:29But they're also learning and they're assessing 5:32and they're thinking about tasks. 5:35And so, while yes, they are just software, 5:39they're also kind of more. 5:41They're somewhere 5:43starting to behave 5:45the ways that humans would behave when they're taking on tasks. 5:48And that's the artificial intelligence piece of this, right? 5:51So, so the question is: are agents, are they just software? 5:55Is that just, are we just going to look at them like traditional 5:57non-human identities and applications and like that? 6:01all it is right. 6:03So that's the first big question I have. 6:06The second question then is: should we 6:13or do we recognize 6:19agents as coworkers? 6:26Now this is actually a point of a lot of debate. 6:31There was there's a good use case from a year or so ago 6:35where an organization actually treated them 6:37from an HR perspective 6:40as an agent was just another worker. 6:43And there was actually, you know, some documented things about that 6:47that didn't really work out so well from a human, 6:50you know, from an HR perspective. 6:52But this is not that. 6:53This is not the question I'm asking here. 6:55This is if a, if an agent is performing 6:58tasks for an organization, for a business unit, 7:02should we treat them like they are just another coworker? 7:07And part of the reason I ask that is 7:09if we think about a case where we have 7:12agents that are supplementing or extending, 7:15virtually extending a support team. And 7:17they're going to pick up and do tasks similar, 7:21or maybe even the same way that a human would process, 7:24support and, and find answers and report on those answers. 7:28If we're virtually treating them like a workforce, 7:31do we recognize them as a coworker? 7:34So this is the next question that I posed to everyone is: 7:37do we recognize them as a coworker? 7:39The third question that I really want to ask then 7:42is, and all of these are very much related, 7:45but should an agent, 7:48do we put an agent 7:52into the directory? 7:57And you know, this can be 7:58your Active Directory, could be an enterprise directory. 8:01It could be a lot of things. 8:02And the reason I posed this question 8:04as in one of my earlier videos, we 8:06were talking a lot about agent identities and governance. 8:11And one of the comments 8:13actually came and said: Please don't tell me that 8:15we're now going to put agents in the Active Directory. 8:17That's actually the question. 8:19Should we be doing that? 8:21Should we be putting, if they're coworkers, 8:24if we think of them as supplementing our workforce 8:27because they're performing tests similar to the way we do it? 8:31Should they be in the directory? 8:33So I posed that question out to everyone. 8:35The next question I'd like to ask 8:38is: are agents, 8:41are they persistent? 8:47or ephemeral? 8:50In other words, 8:54Me, as a human worker, I have a physical existence, right? 8:58So I'm persistent. I come to work, 9:00I am there, 9:02and I perform the tasks that are coming my way. 9:05Agents can actually be, back to the first question 9:09if we, you know, they are software. 9:11So we can actually, we can provision them. 9:14We can de-provision them. We can spin them up, or we can spin them down. 9:17So the question is: should an agent be persistent? 9:20Should it always be waiting around 9:22to take on a task? Or 9:24should it just be brought up to do that task, and 9:29then, we bring it down? 9:31And there's actually some cost implications with this. 9:34There is a cost associated from an IT 9:36perspective of running agents. You 9:39know, if they're running in a container or a part or whatever, 9:41however you're doing that, the question is, 9:43there's a cost associated with that CPU usage, 9:46you know, network traffic, data consumption, whatever that is. 9:50So if they're persistent, they could then 9:53be continually consuming resources. 9:55And do we want that or not? 9:56So this, this is the next question I pose. 9:59The final question I want to pose to everyone 10:03is: does our current IGA system, 10:09our identity, governance and administration. 10:13Does it do what we need? 10:18In other words, is, 10:19is it, is it enough? 10:22And the reason I say that is that we think about 10:25if agents could be ephemeral, if there are supplementing workforces, 10:28if we need to spin them up, we could actually make 10:31thousands and thousands of these agents. 10:33There's actually multiplier 10:35numbers out there now that are all over the place. 10:38Whether you need 5x agents to perform tasks, 10:42if they're. There's a number 10:44that says that if how many people you have in your organization, 10:46you're ultimately someday going to have 45x that in agents, whatever, 10:50however you do your multiplier, 10:52we know there's going to be a lot of agents. 10:55And that means from a governance perspective, 10:57a lot of approvals, a lot of entitlements, 10:59a lot of annual validations. 11:01And if we think about the number of potential agents that should come out, the 11:06question is: will our current way that we handle identities, 11:12really around human identities? 11:14And if we're drawing some parallels here, 11:16is that enough? 11:18So these are the questions. 11:19These are the questions I am posing to everyone. 11:22And what I really would like you to do 11:24is, as mentioned, join in the conversation. 11:28some comments in. 11:29Tell me what you think about one or several 11:31or all of these questions. 11:33Other people can chime in and let's have a conversation. 11:37I do not want this to be a debate. 11:39It's not about right or wrong. 11:40It's not about whether it is or is not. 11:43It's really a conversation about where 11:45we think things are heading. 11:48So please engage and I look forward to reading your comments.