“We believe today that there needs to be a human in the loop for every agent. And that human ultimately, ethically, morally, is responsible for that agent.”
Tomorrow, Today · October 2026
Context
Asked how identity can work for agents that behave probabilistically and can spawn other agents, he answers with a human-responsibility rule for every agent.
Transcript
see we're we're not governing or or or um trying to control the quality of the output right what we're trying to do is making sure that um people have access to the right things in the right environment so so for instance um
…he knows the output, he doesn't know like what the transformer technology really does. Sorry. So, how do you solve the identity problem of an agent which assumes the properties in a very probabilistic way? Yeah, and then agents can spawn agents, and so how do you manage and control all of those?
Well, look, here's what I would tell you: we put things in place foundationally, and we believe today, yeah, today, um, that there needs to be a human in the loop for every agent, right? And that's that human ultimately, ethically, morally, is responsible for that agent, yeah, right? And so, my prediction-is
it the authorizing person, or is it the model provider? Is it the guy who actually trained the model? Is it the guy who gave the knowledge to the model, or is it the guy who actually provisioned the agent? Who's responsible?
Yeah, I think at the end of the day, the human that's responsible for that agent is ultimately responsible. Now, you're going to get into a bunch of things here like policy, right? There's a policy engine, there's a risk engine, there's an intent engine, right? All of these things are going to play…
The behavior didn't come from the person.
More from Tomorrow, Today
