Federal chatbot and virtual-assistant case studies (Digital.gov)
Three Digital.gov articles on real federal chatbot deployments: a cross-agency panel write-up covering USA.gov’s scam-reporting chatbot and the TSA’s AskTSA social-media bot, and two articles from Federal Student Aid on their virtual assistant Aidan — its initial launch and a two-years-later retrospective — covering how each deployment discloses and identifies itself as an AI system to the people talking to it.
License: Public domain — U.S. federal government work (17 U.S.C. § 105).
Key points
- Be upfront that users are talking to a bot, not a person, regardless of whether the bot is named — a bot with a warm enough persona can be mistaken for a human even without an explicit human name, so the disclosure has to be explicit rather than left to be inferred from the bot’s apparent lack of personality. See AI Content Marking.
- A persistent visual motif can double as an ongoing AI-identity marker, not just a one-time disclosure. Federal Student Aid’s virtual assistant Aidan uses an owl icon shown throughout every interaction, chosen partly because it doubles as a standing reminder that the customer is talking to a virtual assistant rather than a live representative. See AI Content Marking.
- When the system can’t answer, hand off to a named escape valve — either a live agent or a self-service knowledge base — rather than leaving the person stuck with an assistant that has run out of answers.