A series by Witnium
The Evidence Question
A short series on LinkedIn about one question: when something goes wrong, can an organisation prove what its AI systems and digital processes actually did? We look at new rules, real cases and everyday practice, and separate what the law requires from what makes a record hold up.
How it is written
The Evidence Question is written with AI assistance and reviewed and published by people. Every post is sealed with WitniumChain.
Every post is sealed
When a post is approved, its exact published text is sealed with a digital witness on WitniumChain. The first comment under each post links to that seal, so anyone can check that the text has not changed since it was approved. Corrections are sealed as new versions; nothing is overwritten.
How the text is normalised before sealing: What is sealed is the published text: the post with the number of days filled in for the day of publication and without @ before tagged names, followed by a blank line and the hashtags separated by spaces. Before sealing, the text is trimmed, all line breaks are converted to one kind, and double spaces are removed.