Palo Alto Networks CEO says enterprise AI adoption lags consumers by years, coding assistants only

Enterprise AI Adoption Trails Consumer Use by Years

Palo Alto Networks CEO Nikesh Arora delivered a sobering assessment of enterprise AI adoption during the company's Q2 earnings call, stating that businesses lag consumer uptake by at least two to three years1

1

. "Consumers are far outstripping enterprise for the moment, but we expect enterprise will surely and slowly get on that bandwagon," Arora explained2

2

. The Palo Alto Networks CEO compared the current business AI trajectory to the cloud computing shift, which took several years before enterprises began migrating applications at scale1

1

.

When pressed about widespread enterprise AI applications, Nikesh Arora could identify only one category seeing significant traction: coding assistants. "Tell me how many enterprise AI apps are you using which are driving tremendous amounts of throughput. I can't think of anything but coding apps," he stated1

1

. This limited deployment presents both challenges and opportunities for cybersecurity vendors like Palo Alto Networks, as coding assistants generate minimal network traffic that security tools can monitor and protect1

1

.

Securing AI Traffic Emerges as Critical Challenge

While enterprise AI adoption remains nascent, Arora noted emerging patterns where customers run millions of tokens through specific applications built with LLM providers1

1

. The challenge isn't network capacity—existing infrastructure handles the load—but rather consolidating and monitoring AI traffic. "How do you get all the AI traffic to be in one place? So you can understand it, provide visibility, look at the ability to control it and be able to act on it," Arora explained1

1

. This growing traffic "needs a different set of controls and tools," driving Palo Alto Networks' AI acquisition strategy2

2

.

The company announced its acquisition of Koi, a startup specializing in agentic endpoint security, to address this evolving landscape4

4

. Koi's capabilities will integrate into Prisma AIRS, Palo Alto Networks' AI security platform, which has already attracted more than 100 customers—a faster scaling trajectory than the company experienced with cloud security4

4

. The acquisition complements the recently completed $25 billion CyberArk deal, which Arora described as essential for securing AI agents operating at machine speed. "We bought CyberArk because when AI agents start logging in at machine speed, logging in becomes the primary attack vector," he said4

4

.

AI Impact on Cybersecurity: Opportunity, Not Threat

Addressing investor concerns about AI replacing security tools, Arora firmly rejected the notion that LLMs pose an existential threat to cybersecurity vendors. "I'm still confused why the market is treating AI as a threat" to the industry, he stated during the earnings call4

4

. He emphasized that LLMs lack the accuracy required for critical security operations, noting they must reach 99.9-percent accuracy before becoming viable replacements4

4

. Furthermore, cybersecurity platforms benefit from proprietary domain-specific data based on real-world threats that LLMs cannot replicate. "To the extent we are creating proprietary data in security, that is not going to be replaced by an LLM," Arora explained4

4

.

Instead of viewing AI as competition, Arora sees it expanding the attack surface and increasing demand for protection. "AI increases complexity... and that increases the need for security," he told the Economic Times3

3

. Most of Palo Alto Networks' business revolves around inspection software—hardware firewalls, software, remote user protection, and cloud security—capabilities that remain essential regardless of AI advancement.

Jobs and Deterministic Problems: Tempering AI Expectations

Arora also addressed widespread fears about AI eliminating jobs, calling such predictions exaggerated. "I don't think 80% of the jobs are going anywhere soon," he stated, explaining that job displacement requires eliminating the need for human judgment—a threshold rarely met in practice. While AI excels at deterministic problems like resolving customer support issues, real-world jobs involve edge cases requiring continuous retraining and human oversight. At Palo Alto Networks, AI drives efficiency and automation, but the company continues hiring more people rather than eliminating roles.

Despite the measured outlook on consumer vs enterprise AI adoption, Palo Alto Networks delivered solid financial results with $2.6 billion in Q2 revenue, representing 15 percent year-over-year growth1

1

. The company reported 23 percent growth in remaining performance obligations, now standing at $16 billion, and projected Q3 revenue would grow at least 28 percent to between $2.941 billion and $2.945 billion1

1

. However, investors knocked six percent off the share price, possibly due to predictions of easing profit margins1

1

. Arora remains confident that as enterprises consolidate their security tools to prepare for AI, Palo Alto Networks' platform approach positions the company to capture demand when enterprise AI adoption inevitably accelerates1

1

.