Margaret Atwood, the acclaimed author of The Handmaid's Tale, delivered a pointed critique of artificial intelligence during an interview at the Babell Literary and Cultural Festival in Porto, Portugal. She recounted her sole experience with an AI chatbot, Anthropic's Claude, which ended in frustration when it provided incorrect information about the British detective series Father Brown.
Atwood characterized the problem as a fundamental flaw: 'garbage in, garbage out.' The chatbot, she explained, gave a wrong answer without realizing it was lying because it lacks human consciousness. 'It's not a human being; it's a large language model,' she told the festival audience, according to a recap by Deadline.
The author's remarks underscore growing unease among cultural figures about AI's reliability in conveying accurate information. Atwood described how Claude 'skimmed a' source and fabricated details, highlighting a persistent issue with AI models that prioritize plausible-sounding responses over factual correctness.
Her criticism resonates as AI tools face increasing scrutiny over hallucinations—confident but false outputs. Atwood's comments suggest that for literary research at least, these systems remain untrustworthy. She called for greater awareness of AI's limitations, emphasizing that machines cannot truly understand or verify the information they generate.
The festival appearance came during a broader tour for Atwood, who continues to engage with contemporary debates. Her blunt assessment may amplify calls for more rigorous factuality in AI systems, though developers like Anthropic are actively working to reduce hallucinations.