Palo Alto Networks CEO warns AI 'Gold Rush' risks trust, security crisis if guardrails are ignored

New Delhi [India], February 19 (ANI): The rapid acceleration of Artificial Intelligence (AI) has turned into a "gold rush" that risks triggering a far-reaching trust and security crisis if companies continue to prioritise speed, scale and competitive advantage over testing, safeguards and responsible deployment, Palo Alto Networks Chief Executive Officer Nikesh Arora said on Thursday.
"There is a risk that the pace in which AI is going, we are going to have a trust and security problem, because there is no thought being given to deploying it responsibly, because it's a gold rush," he said during the AI Impact Summit in New Delhi, themed "Diffusing AI: Human Capital, Infrastructure and Trust."
He further said "Every time I stop to create security controls, or any sort of guardrails, I'm giving up competitive advantage, and time is a competitive advantage right now."
He cautioned that failure to adequately test products particularly those used by children could provoke political and regulatory backlash.
"If they don't think this time around about testing their products first before they launch them, particularly on children, then we are heading for a real trust problem in the future," he said. "You will end up with bans and regulatory policy political reactions that are actually probably not in our long-term interest."
Arora said the bulk of current AI computing power is being consumed by consumer applications. "Eighty percent of AI consumption is in the consumer space," he said, referring to users interacting with chatbots and generating content. "That's where 80% of the compute is going... and about 20%, 30%, give or take, enterprise applications."
While consumers can tolerate imperfections, businesses cannot, he said. "Consumers need 80%, 85% accuracy and they're fine with it. Enterprises need 100% accuracy. If it gives one wrong answer, my business is impacted." He added that safety-critical systems such as autonomous vehicles cannot rely on "an 85% hallucinating model."
On concerns that AI could rapidly displace workers, Arora urged against short-term alarm. "We have to pause before we get paranoid about the fact that human capital is going to be impacted so quickly. I think we have enough time," he said. Technological disruption historically overestimates short-term effects and underestimates long-term transformation, he added.
Rather than reducing demand for labour, AI could sharply increase the need for technical skills. "We're gonna need three to five times more developers in the world than we have today," Arora said, arguing that every intelligent system from industrial automation to robotics requires training, coding and ongoing oversight. Upskilling, he said, will involve a degree of "Darwinism," with engineers independently learning to use tools such as advanced coding models.
Addressing India's startup ecosystem, Arora said the country has made significant strides over the past decade but must move beyond replicating existing business models. "The most successful companies in the world are the ones which have no competition," he said, urging founders to tackle harder, globally scalable problems rather than operate within capital constraints or incremental optimisation.
India's competitive environment, shaped by scarcity and intense competition for opportunity, can both drive resilience and limit ambition, he noted. Solving bigger challenges may require greater risk-taking and capital access.
As AI investment accelerates and data centre infrastructure expands globally, Arora said the central challenge remains balancing innovation with trust. Without adequate guardrails, he warned, the current AI boom could undermine public confidence and invite regulatory responses that may ultimately slow the industry's long-term progress. (ANI)