Keystroke and writing-session records
Strength: Moderate, and widely overclaimed
- What it proves
- That a human typed this text, at human speed, over real elapsed time, and that the words were not swapped afterwards.
- What it does not
- It cannot show you composed the words rather than retyping them. Published research shows transcription produces keystroke timing statistically indistinguishable from composition, and timing patterns can be forged outright. Any product claiming to prove authorship from typing behaviour alone is overclaiming.
- How to get it
- Write in a recorder that chains events as you go. It only covers writing done inside it, so it is a going-forward tool rather than a rescue.
- How long it survives
- As long as the chain verifies. Tamper with any event and the rest stops proving anything.
- Why it matters
- Worth having, and worth describing accurately. An overstated claim collapses the moment someone technical reads it.
Check the evidence in your own document
The analyzer reads the process record inside a .docx and reports each signal separately, with its limits stated.
Analyze a document Read the full guideMore in types of evidence
Other signposts