When AI Fabricates Truth: Lessons from a Controversial Interview

The interview with the author of Future of Truth began as a straightforward inquiry into how artificial intelligence reshapes our grasp of reality, but quickly devolved into a confession that the book's quoted passages were largely fabricated by a language model. The revelation that AI-generated quotations had been slipped into a work presented as rigorous scholarship sent shockwaves through academia and journalism. It exposed a fragile epistemic contract: readers trust the author's voice, yet the text itself may be a collage of algorithmic invention, blurring the line between citation and invention. This situation reflects that the credibility of a media outlet depends on the transparency of its sources, and that undisclosed AI use threatens to undermine confidence in a difficult-to-repair way. This episode is not an isolated lapse but the symptom of a broader crisis. As generative models become ubiquitous, the provenance of every statement is called into question, echoing historic scandals of forged documents while adding a technological twist that complicates verification and fuels the temptation of large-scale manipulation. For newsrooms, the incident underscores the need for systematic auditing of texts to detect synthetic passages, a labor-intensive task that demands AI-detection tools and heightened editorial diligence, while raising legal and ethical questions about content responsibility. Looking ahead, the scientific community, editors and regulators must collaborate to establish clear disclosure standards, ensuring that AI serves to illuminate rather than obscure truth and safeguarding a trustworthy information ecosystem for the public. Such measures are essential to preserve democratic discourse and prevent the erosion of shared reality.