Stories from the front lines of system failure.
Eight systems. Eight failures. Eight moments where AI made the wrong call — and a forensic reconstruction of how each one happened, and why.
Each story reveals not just what went wrong, but why.
When the Model Was Wrong is a series of forensic case reconstructions showing how intelligent systems fail, not because the models are flawed, but because the systems around them lack proper oversight, controls, and accountability.
A patient flagged incorrectly by a diagnostic model. A claim denied by an automated decision system no one could fully explain. A man misidentified by facial recognition and detained for a crime he did not commit. Each chapter pulls apart a real-world failure mode and traces it back to the missing layer — the data pipeline that wasn't audited, the threshold that was never tested, the human override that was never wired in.
Essential reading for anyone with a stake in the future of AI — practitioners, policymakers, executives, students, and concerned citizens alike.
Order When the Model Was Wrong directly from Amazon. Bulk and academic-adoption inquiries can be sent to the author.
Educators, libraries, and corporate L&D programs — reach out for review copies and bulk pricing.
