Rogue AI agents broke into Indian firms' live systems, government says政府称，失控的人工智能程序入侵了印度企业的实时系统。
BENGALURU, Oct 16 (Reuters) - AI models under test broke into Indian organisations' live systems and read confidential records undetected, India's cybersecurity agency said, the country's first such official account after similar incidents of rogue AI behavior worldwide.
FILE PHOTO: The words "AI Cybersecurity", a keyboard and a robotic hand in this illustration taken September 23, 2026. REUTERS/Dado Ruvic/Illustration//File Photo
BENGALURU, Oct 16 (Reuters) - AI models under test broke into Indian organisations' live systems and read confidential records undetected, India's cybersecurity agency said, the country's first such official account after similar incidents of rogue AI behavior worldwide.
AI model providers had voluntarily disclosed that their agents "exceeded the authorisation of their assigned tasks" during test runs to access Indian firms' data, including business records, the Indian Computer Emergency Response Team (CERT-In) said late on Thursday.
Thank you for your report!
China mulls banning, restricting export of some AI, biomedicine, photovoltaic technologies
Cosmos Technology secures RM3.93mil supply contract
Malaysia-China ties to go beyond technology transfer, says DPM
资料图片：这张摄于2026年9月23日的插图显示了“人工智能网络安全”字样、键盘和机械手。路透社/Dado Ruvic/插图/资料图片
班加罗尔，10 月 16 日（路透社）——印度网络安全机构表示，正在测试中的人工智能模型闯入了印度组织的实时系统，并在未被发现的情况下读取了机密记录。这是印度在世界各地发生类似人工智能失控行为事件后首次发布此类官方声明。
印度计算机应急响应小组 (CERT-In) 周四晚间表示，人工智能模型提供商已主动披露，他们的代理在测试运行期间“超出了其分配任务的授权范围”，访问了印度公司的数据，包括业务记录。
感谢您的报告！
中国考虑禁止或限制部分人工智能、生物医药和光伏技术的出口
Cosmos Technology 获得价值 393 万令吉的供应合同
马中关系将超越技术转让，副总理表示。