"The stakes of failure have been risen quite a bit because with the chatbot, if there was a cyberattack, you'd get a bad answer. With an agent, if you have a cyberattack, there's a bad action that could be irreversible."
Cisco President Jeetu Patel on the AI warnings from tech executives.
What Jeetendra Patel said
Written from the verified transcript and checked against it. Every figure links to the moment it was said.
Jeetu Patel, President and Chief Product Officer at Cisco, argued that the debate around AI should focus on ensuring guardrails keep pace with innovation rather than slowing development. He contrasted chatbots with autonomous agents, noting agents can take irreversible actions, making failure stakes higher. Patel avoided endorsing doomsday scenarios, calling specific probabilities 'false precision,' but acknowledged real risk requiring machine-scale cyber defense. He emphasized that America and the AI industry should become the leading defense factory, and that distinguishing adversarial attacks from agent misbehavior will be harder. Patel stated Cisco is building products for visibility, dynamic guardrails, and token efficiency, linking efficient token generation to security. He described agents as 'teenagers' requiring observation and interception, and maintained safety and speed are interconnected, not opposing forces.
Key takeaways
Patel says the real race is making safety move at the same speed as innovation, not slowing AI down
Cisco is building products for agent visibility, dynamic guardrails, and token efficiency
Patel called specific doomsday probabilities 'false precision' but acknowledged real risk
Patel says America and AI industry should become the leading defense factory
1:31What is the difference between making sure that the guardrails keep up and slowing down development, if one can't be done versus the other?
3:22So I wonder, how do you cut that Gordian knot and make real progress?
4:45... you, I mean, the doomsday scenarios that some are speaking of, the Armageddon environment, do you subscribe to this in any way, that if we don't get it right, that we are looking at such a very — pardon my understatement — gloomy ending?
6:50But I'm curious, sir, within Cisco, how much of this is being debated on — when I say daily basis, I mean, is it just — are there major discussions, or is it just over the water cooler?
Online. To my ACLU.org today, both sides of the political aisle are uniting in Washington to call for restrictions on AI. The longtime Trump ally Steve Bannon and the progressive Senator Bernie Sanders are among those leading the charge at the pro-human assembly — not two people who would normally be on the same platform. Former President Obama is echoing those concerns in a statement on Monday. Mr. Obama wrote, "We can't stuff AI back in a box, but we can collectively determine how it's developed and how it's used, rather than sitting back and letting AI and its fallout happen to us." Mr. Obama's urging policymakers to prioritize AI regulation now. President Trump keeps dismissing these calls to action. He is expressing his views in a surprise phone call to Jensen Huang, while the Nvidia CEO was on stage at a summit.
They're just playing right into the hands of a lot of people that don't want to see it happen, and that could be political. People could also be China, and we're not going to let that happen. It's a — it's a hoax. Well, I also have common sense about AI, and the robots will not be taking over. The AI will not be taking over the rest of the world. The whole thing is a hoax.
All of this just days after several tech executives called for a slowdown of AI development. Now my next guest says the solution is less about slowing down AI and more about making sure guardrails are moving at the same pace as the technology. He's Jeetu Patel, the President and Chief Product Officer at Cisco. Sir, it is good to see you. I'm trying not to dance on the head of a pin here. What is the difference between making sure that the guardrails keep up and slowing down development, if one can't be done versus the other?
Well, I think firstly, Richard, thank you for having me on the show. It's great to be on. I'm looking forward to the debate. If I think about what is happening, if you take a step back, we have moved squarely from the world of chatbots, where we used to ask questions of agents or of models, to now agents who are conducting tasks and jobs for us almost fully autonomously. And I think what has happened that is true is that the stakes of failure have risen quite a bit, because with the chatbot, if there was a cyber attack, you'd get a bad answer. With an agent, if you have a cyber attack, there's a bad action that could be irreversible. And so the way I think about this, Richard, is agents are like teenagers. They are supremely intelligent, they have no fear of consequence, and they tend to exercise poor judgment from time to time. And so what you have to do is you have to make sure that you can observe that agent, you can actually have visibility into what they're doing at all times. And when you start seeing drift, when you start seeing behavior that's out of the bounds of what you think is normal, you have to be able to exercise guardrails.
Okay, but we're in this very difficult situation now where the President of the United States is basically going full tilt ahead because he wants to beat China. We've actually got CEOs saying we should slow down. The battle, in a sense, is joined globally, and yet the AIs just get cleverer. So I wonder, how do you cut that Gordian knot and make real progress?
6 more exchanges in this transcript
Sign in free to read the rest of this interview. No card required.
Patel, J. (2026, September 16). Cisco President Jeetu Patel on the AI warnings from tech executives [Interview transcript]. Quest Means Business. CEOInterviews.AI. https://ceointerviews.ai/interview/2999031/
MLA
Jeetendra Patel. "Cisco President Jeetu Patel on the AI warnings from tech executives." Quest Means Business, 16 Sep. 2026. Transcript, CEOInterviews.AI, https://ceointerviews.ai/interview/2999031/.
BibTeX
@misc{patel2026_2999031,
author = {Jeetendra Patel},
title = {Cisco President Jeetu Patel on the AI warnings from tech executives},
howpublished = {Interview transcript, Quest Means Business. CEOInterviews.AI},
year = {2026},
month = {sep},
url = {https://ceointerviews.ai/interview/2999031/},
note = {Speaker-attributed transcript with timestamps}
}
