Margaret Atwood told an interviewer at the Babell Literary and Cultural Festival in Porto that her single experience with an AI chatbot ended with the model giving her incorrect information about the British detective series Father Brown. The author, who wrote The Handmaid’s Tale and The Blind Assassin, was speaking to Deadline when she described using Anthropic‘s Claude to find details on the show. Atwood noted that the system provided a wrong answer or lied without realising it was doing so because it is a large language model rather than a human being. She stopped the quote mid-explanation after noting the model had skimmed a source.
This interaction highlights a core limitation of current generative systems where accuracy depends entirely on the quality of training data and the specific prompts given. Atwood’s point suggests that without human oversight, these tools cannot be trusted for factual retrieval. The issue is not that the technology is incapable of learning but that the output reflects the flaws in the input data it processes. This reality means writers and researchers must verify every claim made by such software before using it in serious work.
