Palo Alto CEO says AI isn’t great for business, yet

Palo Alto Networks sees AI delay

Palo Alto Networks CEO Nikesh Arora told investors on Tuesday that enterprise AI adoption currently lags behind consumer use by at least two years. During the company’s Q2 earnings call, Arora noted that while individuals have embraced generative AI tools, businesses are moving much slower to integrate the technology into their core operations.

Arora compared the current AI trajectory to the early days of cloud computing. He reminded analysts that it took two to three years for enterprises to begin migrating serious applications to the cloud after the initial hype began. Right now, most businesses are still in the experimentation phase rather than the deployment phase.

The CEO challenged the idea that AI is already driving massive enterprise utility. He asked the call participants to name enterprise AI applications that are currently driving significant network throughput. Arora concluded that, for now, coding assistants are the only category seeing meaningful adoption in the corporate world.

Coding tools don't drive revenue

The rise of AI-powered coding assistants provides little immediate benefit to Palo Alto Networks. These tools do not generate the type of heavy network traffic that requires advanced security monitoring or firewall intervention. Because the data stays relatively contained, Palo Alto cannot easily apply its security stack to these specific interactions.

Arora believes his competitors in the security industry are facing the same reality. He described the current market as an arms race where vendors are rushing to build AI security platforms before the demand fully arrives. The goal is to have the infrastructure ready by the time enterprises move past coding help and into more complex AI use cases.

Despite the slow start, Palo Alto is seeing some early signs of high-volume AI usage. A small number of customers are now running millions of tokens through specific applications built with large language model (LLM) providers. This activity is finally showing up as visible traffic on local area networks (LANs).

The challenge of consolidating traffic

Existing enterprise networks are not yet struggling to handle the volume of AI-related data. However, Arora identified a different problem for security teams: traffic consolidation. Most AI traffic is currently fragmented, making it difficult for security officers to monitor or control how data moves between employees and LLMs.

Security teams currently lack a single point of visibility for these new AI workflows. Palo Alto is focusing on tools that bring all AI-related traffic into one place. This allows the company to provide visibility, apply controls, and act on potential threats before they compromise corporate data.

As these traffic patterns grow, they will require a fundamentally different set of tools than traditional web traffic. Palo Alto is positioning its platformization strategy to address this. The company wants to replace the "tangle" of disconnected security tools with a unified platform that handles both traditional and AI-specific threats.

Palo Alto acquires agentic AI startup

Palo Alto Networks officially confirmed the acquisition of Koi, a startup specializing in agentic AI endpoint security. The deal ends months of rumors regarding the company’s interest in the firm. Koi’s technology focuses on securing autonomous AI agents that can perform tasks on behalf of users.

This acquisition is part of a broader shopping spree to bolster Palo Alto’s AI security portfolio. Arora highlighted the recent additions of Chronosphere and CyberArk as evidence of this strategy. These companies provide critical pieces of the puzzle for securing the future AI-driven enterprise.

• Koi: Focuses on securing autonomous agents and endpoint AI interactions.
• Chronosphere: Provides cloud-native observability to track complex data flows.
• CyberArk: Strengthens identity management, which is critical as AI agents gain more permissions.

Arora expressed confidence that Palo Alto now owns the necessary products to secure the next wave of corporate tech. He argued that customers are already consolidating their security spend onto Palo Alto’s platform to prepare for the eventual AI shift. This consolidation is a core part of the company's long-term growth plan.

Revenue grows but stock falls

Palo Alto Networks reported $2.6 billion in revenue for the second quarter, a 15 percent increase compared to the same period last year. The company’s subscription business performed even better, showing significant momentum in long-term contracts. Remaining performance obligations (RPO) grew 23 percent to reach $16 billion.

The company also issued strong guidance for the upcoming third quarter. Executives expect revenue to land between $2.941 billion and $2.945 billion. This would represent a year-over-year growth rate of at least 28 percent, signaling an acceleration in sales.

The financial breakdown for the quarter included several key metrics:

• Total Revenue: $2.6 billion (15% YoY growth).
• Remaining Performance Obligations: $16 billion (23% YoY growth).
• Q3 Revenue Guidance: $2.941 billion to $2.945 billion.
• Billings Growth: Maintained steady pace despite the shift to "platformization."

Investors were not satisfied with the report despite the revenue beat. Shares of Palo Alto Networks fell 6 percent in after-hours trading following the announcement. The decline appears linked to forecasts suggesting that profit margins may ease as the company continues its aggressive acquisition and platform-building strategy.

The long road to AI security

Palo Alto is betting that the current lull in enterprise AI adoption is temporary. Arora’s strategy relies on the belief that once businesses figure out how to use LLMs safely, they will need a massive security overhaul. By building the platform now, Palo Alto hopes to capture that spend before competitors can react.

The "platformization" strategy involves convincing customers to dump their existing point-solution vendors in favor of Palo Alto’s full suite. In some cases, the company has offered free or discounted services to help customers make the switch. This tactic is designed to build the $16 billion backlog that the company reported on Tuesday.

While the market remains skeptical of the short-term profit outlook, Palo Alto is doubling down on its infrastructure. The company believes that securing the AI era requires more than just firewalls; it requires a total grip on how data moves within the enterprise. For now, the company is waiting for the rest of the business world to catch up to the AI hype.