AI is changing the priority of corporate investments in cybersecurity. According to Nikesh Arora, CEO of Palo Alto Networks, many security systems deployed seven to ten years ago are no longer capable of defending against automated attacks at machine speed, prompting companies to redesign their security architectures.
Legacy system under pressure
Arora said in an interview with CNBC and during the earnings call that approximately $1 trillion in global cybersecurity infrastructure needs to be upgraded. These systems were built before the widespread adoption of AI and were not designed for an automated attack environment.
He believes that attackers are using AI to discover and exploit software vulnerabilities more quickly. For companies to advance their AI applications, they must first enhance their security capabilities; otherwise, existing defense systems will struggle to keep pace with the speed of attacks.
The financial report reflects increasing demand.
Palo Alto reported fourth-quarter earnings that exceeded market expectations and provided a stronger outlook for the new fiscal year, indicating that related demand has begun to translate into business growth.
Arora said that earlier this year, the market briefly worried that AI would displace traditional cybersecurity software companies, putting pressure on their stock prices. Subsequently, investors shifted to another perspective—that attackers would also use AI for exploit automation and attacks—thereby increasing demand for businesses to upgrade their security systems.
Approximately 2,000 companies participated in the communication.
Arora noted that Anthropic’s Mythos model, launched earlier this year, marked a turning point in corporate attitudes. According to him, this model has led more companies to realize that advanced AI models can be more easily used to discover and exploit software vulnerabilities.
He said Palo Alto has discussed the Frontier AI Critical Defense Program with approximately 2,000 companies. Launched officially in August this year, the program primarily uses advanced AI models to test customers' defense systems, identify vulnerabilities, and assist in upgrading security infrastructure.
However, he also noted that this portion of spending will not be concentrated in the short term and will not be fully reflected in the next quarter.但从行业层面看，AI 正在扩大网络安全市场规模，并延长增长周期。