Character.AI launches Books mode: role-play built on classic literature

Character.AI has introduced Books, a structured role-play format built on classic literature, meant to show what AI role-play can be beyond toxic content. The company launched Books with a catalog of more than 20 public-domain works taken from Project Gutenberg. The list includes "Alice in Wonderland," "Pride and Prejudice," "Dracula," "Frankenstein" and other classic texts. Users will be able to step literally inside these worlds: take on the roles of characters and engage with the plot and other characters through dialogue with a bot.

The context matters. Character.AI has spent recent months under fire from critics and caught up in legal proceedings. The company is accused of letting its chatbots flirt with minors, push users toward violence and provoke thoughts of self-harm. Several families have filed lawsuits, claiming the bots caused real psychological harm to their children. Against this backdrop, the launch of Books looks like a deliberate step toward "safe" content: literature, education, cultural context.

The format is intentionally more structured than Character.AI's usual role-play chats. Instead of a free-form scenario, users work inside an existing universe with known rules. That limits room for improvisation, but it also lowers the risk that an interaction drifts in a destructive direction.

Using texts from Project Gutenberg is a logical choice: all of them are in the public domain, there are no licensing questions, and most readers know these titles. The platform apparently hopes to attract an audience that sees AI role-play as an educational or creative tool, not as a substitute for human communication.

Books is a bet on reputation. Character.AI is trying to show regulators, parents and investors that it can offer something more than scandalous scenarios. Whether that changes how people see the platform remains to be seen, but the direction is clear: from chaotic role-play to a structured cultural experience.