When Documents in Evidence No Longer Reflect Human Thought
One of the assumptions underpinning modern commercial disputes is that documentary evidence is generally regarded as being more reliable than human memory.
AI may begin to unsettle that assumption.
Increasingly, commercial decisions are being made in environments shaped by AI-generated summaries, AI-assisted drafting, autogenerated meeting notes and synthesised reporting. In many businesses, people are beginning to rely less on primary material and more on AI-mediated versions of it.
Historically, documents carried evidential weight because they were assumed to reflect contemporaneous human thought — what somebody knew, intended or understood at a particular point in time. But what happens when the email was AI-assisted, the meeting note was autogenerated, the recommendation was machine-produced, or the underlying material was never fully reviewed by a human in the first place?
At that point, the dispute is no longer only about what the document says. It is also about authorship, attribution and whether the document genuinely reflects human reasoning at all.
AI may therefore create a paradox for commercial disputes: businesses may end up with more documentary material than ever before, but less certainty about whether those documents genuinely reflect contemporaneous human judgement.
If documentary records become increasingly AI-mediated, courts and tribunals may eventually need to confront a more fundamental evidential question: when documents no longer reliably reflect human reasoning, whose account of events should ultimately be trusted?