AI on Trial by Kai London
An algorithm just made a decision that changes someone's life. Could you prove, in a courtroom, that it was fair? AI on Trial is the practitioner's field guide to governing AI so that every consequential decision is explainable, auditable, and defensible — before the machine, not the institution, ends up on trial.
By the end of this book you will be able to:
Apply the Evidence Chain Model™
Take any AI output and say, with confidence, whether it could survive scrutiny in court or before a regulator.
Run the SCALES™ framework
Sovereignty, Compliance, Architecture, Legacy, Ethics, Security — find where your AI governance breaks before someone else does.
Decide admissibility
Build the audit trail, model documentation, and human-oversight record that make AI-generated material defensible.
Answer the board's liability questions
What the AI did, who is accountable, what it would cost, which control held it, and where the proof lives.
Map the "£1 Million AI Mistake"
Turn a single AI failure into a funded, staged remediation plan tied to real exposure.
Translate regulation into practice
Mapped to the EU AI Act, GDPR, NIS2, DORA, the Council of Europe AI Convention and the NIST AI RMF.
Grounded in real authorities — from Mata v. Avianca and State v. Loomis to the Daubert standard — with a full Table of Cases, Table of Statutes & Instruments, and a sourced verification note. Written for Chief Justices and judges, General Counsel and barristers, Chief AI and Chief Risk Officers, compliance and procurement leaders, board members, vendors, academics and policymakers.
Before AI judges us, we must judge AI. This is how.
About the author
Professor Kai London — CISSP, CISM.
An internationally recognised cybersecurity executive, board advisor and Founder & CEO of Quantum AI Systems Security LLC, writing at the convergence of AI, governance and operational resilience. Honorary Professor and Researcher at UCL.