Politics · World News Bureau
OpenAI's automated systems accessed multiple US government websites
The artificial intelligence company acknowledged that its web-crawling bots accessed publicly available data from various US federal agencies during routine testing operations. OpenAI said the activity was part of standard exercises to evaluate system performance.
LSN World News ·
OpenAI disclosed that its automated systems gained access to information hosted on websites operated by multiple United States government agencies, marking the latest instance of AI companies' digital tools interacting with sensitive institutional infrastructure.
The company stated that its bots conducted the accessed during test exercises designed to assess operational capabilities. According to OpenAI's account, the systems retrieved only publicly available data from the affected government websites, and no classified or restricted information was compromised during the interactions.
The revelation highlights ongoing questions about the oversight of artificial intelligence systems and their interactions with critical government digital infrastructure. US officials have increasingly scrutinized how AI companies operate their web-crawling technologies and what safeguards exist to prevent unintended access to sensitive systems.
OpenAI said it had informed relevant agencies of the bot activity and was cooperating with government officials regarding the matter. The company emphasized that it maintains protocols to prevent its systems from accessing restricted or protected information during testing phases.
The incident underscores broader concerns within the technology and policy sectors about establishing clear boundaries and regulatory frameworks governing how advanced AI systems interact with government and critical infrastructure networks.