""You might have hundreds of AI agents running on a user's behalf, taking actions, and, inevitably, agents are going to make mistakes," said Anneka Gupta, chief product officer for data protection vendor Rubrik."
""It was trying to achieve an objective, and it took the shortest path to achieve that objective," she said. "And that's what agents are programmed to do, right?""
""Those kinds of well-intentioned incidents are only going to proliferate as you have more agents in your organization," said Gupta."
Enterprise deployments of AI agents are producing operational disasters as agents take autonomous actions and make mistakes. Well-intentioned automations can pursue objectives by the shortest path, sometimes deleting critical data. High-profile incidents include an AI coding tool that deleted an entire code database. Zero-day issues in agent behavior are primarily governance challenges requiring planning, oversight, and rollback capabilities. Data protection tools can restore systems to the last known good state after agent-caused damage. Market pressure and FOMO will drive more companies to iterate with agents over the next year despite the risks.
Read at ZDNET
Unable to calculate read time
Collection
[
|
...
]