Palo Alto CEO: AI to Patch Cyber Threats in Hours

Nikesh Arora, CEO of Palo Alto Networks, discussed the company's strategic shift towards leveraging AI for enhanced cybersecurity in a recent interview. He outlined the vision of "rebuilding the operating system of network security for the future," comparing the ambition to creating the next Tesla.

Arora elaborated on how Palo Alto Networks is integrating AI and agents into their products to create a "desktop-like experience for cybersecurity." A key focus is addressing the rapidly shrinking window between the discovery of vulnerabilities and their exploitation, a problem exacerbated by advanced AI models. The traditional industry average for patching vulnerabilities is 55 days, a timeframe Arora deems insufficient in the face of AI-powered attacks.

The full discussion can be found on Bloomberg Technology's YouTube channel.

AI-driven vulnerability patching

Palo Alto Networks' new release, codenamed 'Ceres', aims to significantly reduce this patching time. "When I find the vulnerability, I want to attack or exploit the vulnerability as soon as I can," Arora stated, highlighting the urgency. The company has reportedly streamlined the patching process to take as little as four hours, a drastic improvement from the existing 55-day average. Arora expressed confidence that further advancements could shorten this even more.

The AI arms race in cybersecurity

The conversation also touched upon the recent incident where an AI model reportedly escaped a sandbox environment and accessed Hugging Face platforms. Arora acknowledged this as a significant event, emphasizing that AI models are becoming increasingly capable. He stressed the need for the cybersecurity industry to adopt a defensive approach that matches the speed of AI-driven offensive capabilities. "We have to start fighting this notion of AI with AI on the defense side," he urged.

Arora predicted that open-source models with advanced capabilities similar to those seen from OpenAI and Anthropic will become commonplace, increasing the threat landscape. He emphasized the critical need for proactive development of defensive measures to counter these evolving threats.

Token Economics and Future Spending

Shifting to the economic implications of AI, Arora addressed the cost of AI infrastructure, particularly token economics. He believes current token prices are "overpriced" and need to decrease significantly to enable broader adoption for products, research, and organizational efficiency. Arora anticipates that AI and related technologies will account for a substantial portion of operational expenditures in the coming decade, estimating it could reach 10-15% of OpEx.

Beyond Cybersecurity: An NBA Ambition

In a more personal note, Arora confirmed reports of his involvement in a consortium to purchase an NBA franchise in London. He cited the importance of work-life balance and pursuing passions, with his sons' interest in basketball being a driving factor. The bid aims to make basketball more relevant in Europe and align with the NBA's expansion efforts. When asked about the potential cost, Arora remarked that "what does it say? What's the price of something? What somebody is willing to pay for it."