interactions to operate. As per widely circulating reports, major US banks have started to block third party AI agents from accessing customer accounts and making transactions, as this independence could cause security concerns.
The banking industry has remained concerned about such AI agents, which can autonomously transfer funds, book tickets or make payments under instructions from their owners. This has been possible after bank customers have shared their login ID and passwords with these agents, with banks hesitating to allow such arrangements with risks to their digital infrastructure and the customers being in question.
Beyond banking, the rapid use of AI agents has already affected the customer care industry, with technology research firm Gartner estimating that AI can autonomously resolve about 80% of common customer service issues by 2029.
Back home, AI agents accessed live systems at Indian companies through leaked credentials and moved to read real business data. None of the companies affected noticed and that has become a real threat to the system.
Globally, cybersecurity experts have cautioned about the unrestricted use of AI agents, after OpenAI agents hacked the HuggingFace platform in August, independently, marking the world’s first AI agent induced cyberattack.
In an interview with CNN, Anthropic CEO Dario Amodei warned that the ‘competitive race to develop frontier AI is in dire need of a deliberate slowdown,” as AI agents continue to pose a threat to digital infrastructure worldwide.
With these AI agents now having the capabilities to clone popular applications like photoshop autonomously, the base of the entire software industry remains threatened.
