Palo Alto Networks CEO Nikesh Arora says AI job losses are ‘overstated: We need five times more skilled people

Palo Alto Networks CEO Nikesh Arora has said that the job loss due to AI is “overstated”, at least for now. In an interview with Economic Times, Arora said: “I don’t think 80% of the jobs are going anywhere soon”, adding “We need three to five times more technically skilled people globally than we have today. Everything that will get AI-enabled will have to be done by some technologist”. He told the publication that one major requirement for AI to take over a job is that you don’t need human judgment in that job anymore. “That may be true in certain scenarios. AI works well for deterministic problems…say, resolving customer support issues. But real-world jobs involve edge cases,” Nikesh Arora said. Giving an example of autonomous driving, he said that models constantly need retraining for unusual scenarios. “The same applies across industries. You need structured data, edge-case training, and continuous learning systems. That doesn’t happen overnight,” he added. When asked by threat to software as a sector due to the threat of automation by AI, Nikesh Arora said: “There’s a fear that the software industry is under attack from AI — it will make software easier and faster to build, and therefore reduce the need for traditional software spending. That may be true in some subsectors. If your product is purely analytical that can be recreated easily using AI, there could be disruption. If your system of work can be automated by agents, there is risk.”He also dismissed the idea that cybersecurity could be a collateral damage in the AI wave. Instead, he sees AI expanding the demand for protection. “I’m sure half the speeches being made here are written by some sort of LLM helping scriptwriters. But security works differently, because we inspect traffic. LLMs can tell you a file or an address is bad and we can too. Most of our business revolves around inspection software, whether it’s hardware firewalls, software, remote user protection or cloud security. That’s why I don’t believe cybersecurity is under threat from AI. If anything, AI increases complexity... and that increases the need for security,” he said.