The AI company said the incident "illustrates how influence actors can use AI as a supporting tool within a broader effort to manufacture authority."