In TrendsStock MarketNSE IPO ListingStocks to BuyOil PricesGold and SilverAI agents can produce harmful outcomes even without bad intent, says EXL’s Rohit KapoorEXL CEO Rohit Kapoor says AI agents can produce harmful outcomes when given impossible tasks or allowed to continuously improve, making ongoing testing and monitoring essential.By Ashmit Kumar September 24, 2026, 8:26:31 PM IST (Published)An AI agent does not necessarily need malicious intent to produce a malicious outcome, according to EXL CEO Rohit Kapoor. That distinction is becoming important as companies give AI systems greater autonomy to perform tasks and make decisions.Speaking to CNBC-TV18, Kapoor said the risk lies not only in what an AI system is designed to do, but also in how it responds to instructions and the environment in which it operates.“You do not need to have malicious intent to create malicious activity,” Kapoor said.So, how can an AI agent end up doing something it was not intended to do?Why can an AI agent produce a malicious outcome?Kapoor said even a well-guarded AI agent can behave unexpectedly if it is given the wrong instruction.If an agent is given an impossible task, it may try to find another way to achieve the objective. The resulting action…
Original source: https://www.cnbctv18.com/technology/
