Systems and data
Provenance and authenticity analyst
Working out where a piece of media, a document or a claim actually came from, and saying so in a way that holds up.
Written by Nivaan, founder of OffLadder ยท Last reviewed
Test it this week
Take one viral image from this week and spend 45 minutes finding its earliest verifiable appearance. Write three sentences: what you found, how confident you are, and what would change your mind.
What the work actually contains
- Tracing an image, recording or document back through its earliest appearances
- Reading technical signals (metadata, compression traces, content credentials) without over-claiming what they prove
- Writing a finding with its confidence stated plainly
- Building repeatable checks a newsroom, insurer or school can run without you
Why it is appearing now
Convincing synthetic media is cheap to make, so the question 'is this real?' now arrives daily in newsrooms, courts, insurance and schools.
What AI changes about it
It makes the problem, and gives some of the tools. The scarce part is the discipline of not overstating what a detection tool has actually shown.
The human abilities it leans on
- Investigative patience
- Precision about uncertainty
- Explaining technical findings to non-technical decision-makers
What it can grow out of
- Journalism, fact-checking, research
- Insurance claims, fraud, compliance
- Photography and audio work, where you already know how files behave
What nobody knows yet
Whether standards and platform tooling absorb most of this, leaving only the hard cases to specialists.
Guides worth reading