Artificial intelligence systems increasingly interpret government information outside the environments where that information was originally published. That shift creates a new infrastructure problem. Traditional government publishing systems were designed for human interpretation. A resident visited a website, opened a PDF, followed a social media account, read a press release, or subscribed to emergency alerts. Context existed naturally inside the publishing environment itself. Logos, department pages, timestamps, layouts, and surrounding content helped establish authority. AI systems do not interpret information that way. Instead, they: extract fragments compare sources summarize information reconstruct responses rank competing signals synthesize across disconnected systems As a result, attribution signals can weaken after information leaves its original publishing environment. This is one of the reasons AI Citation Registries are beginning to emerge. What Is an AI Citation Registry?…