Dailyhunt
Palo Alto Networks CEO Warns AI Is Quickly Uncovering Years Of 'Bad Code'- Calls for Urgent 'Kill Switches'

Palo Alto Networks CEO Warns AI Is Quickly Uncovering Years Of 'Bad Code'- Calls for Urgent 'Kill Switches'

Benzinga 3 days ago

Palo Alto Networks Inc.(NASDAQ:PANW) CEO Nikesh Arora issued a stark warning regarding the intersection of generative artificial intelligence (AI) and enterprise cybersecurity, emphasizing that the tech industry faces a rapidly closing window to secure AI agents and patch legacy vulnerabilities.

Exposing ‘Bad Code’ At Unprecedented Speeds

Arora, speaking with the South Park Commons partner Aditya Agarwal, explained that while large language models (LLMs) are becoming proficient at generating code, they are even more adept at analyzing and breaking it. “Humans have been writing bad code for a very long time,” Arora stated, noting that AI will ruthlessly scrutinize open-source, custom, and vendor-deployed software.

This dual-edged capability will trigger a massive influx of exposed flaws. “We will find vulnerabilities which should have taken us 10 years to find in 6 months,” he predicted.

This compressed timeline will severely challenge IT departments, forcing them to scale from handling a few routine patches a week to potentially hundreds, as AI swiftly identifies misconfigurations and uncovers hidden attack paths across enterprise networks.

Preparing ‘Kill Switches’ For Rogue Systems

Compounding this threat is the rushed deployment of AI tools without built-in security frameworks. Arora likened the current AI landscape to the aviation industry before the TSA (Transportation Security Administration), highlighting that security is too often treated as a reactive afterthought.

To combat this, the cybersecurity firm is actively building the necessary scaffolding and agentic gateways to monitor and govern these technologies. Arora emphasized the critical need “to build the kill switches for when AI goes rogue, which it will, it has, and will continue to do so for long.”

He warned that most organizations currently lack fundamental visibility into the safety, security, and potential backdoors of the models operating within their environments.

The Looming Six-Month Window

Despite these imminent threats, Arora remains fundamentally optimistic that the industry will successfully navigate this bottleneck.

Disclaimer: This content was partially produced with the help of AI tools and was reviewed and published by Benzinga editors.

Image via Shutterstock

Dailyhunt
Disclaimer: This content has not been generated, created or edited by Dailyhunt. Publisher: Benzinga