About Lenstrum
How Lenstrum analyzes content
Every analysis breaks into three parts — Framing & Perspectives, AI Involvement, and Source Context — each rated with its own confidence level, never presented as a verdict.
Accuracy & Validation
Lenstrum's analysis engine is validated against a benchmark corpus, tested against AllSides and Ad Fontes Media ratings where a third-party standard exists. We publish our methodology and full accuracy figures openly, because trust in a transparency tool has to be earned, not assumed.
Data sources
Claude (Anthropic) — performs bias, AI detection, and source context analysis via the Anthropic API.
Supadata — retrieves YouTube transcripts for video analysis.
Brave Search — surfaces related coverage from left, centre, and right outlets.
What Lenstrum does not do
Lenstrum does not fact-check. A well-sourced, accurate article can carry an ideological lean — lean is a framing signal, not a quality judgment.
Lenstrum does not rate creators or outlets. It analyzes specific pieces of content, not the broader output of a source.
Lenstrum cannot process content without available text. Videos without captions, paywalled articles with no excerpt, and private posts cannot be analyzed.
Who's behind Lenstrum
Two people can look at the same facts and believe completely different things. Usually nobody is lying. It comes down to how the facts were framed, and by whom, before either person had a chance to think it through.
Staying informed now takes real work: checking several sources, comparing versions of the same story, guessing where the spin is. AI-generated content, much of it built to look credible and spread fast, adds to the pile. It has started to feel like a part-time job.
I'm not a journalist or a media researcher. I'm not a computer scientist either. Before Lenstrum, I worked in law enforcement, where investigations run on something called reasonable and probable grounds: the idea that any reasonable, informed person looking at the same facts should land on roughly the same conclusion, once you strip out assumptions and who's telling the story.
I wanted that same discipline applied to what people read and watch online. A tool that looks at a piece of content the way an investigator would: what's being claimed, how it's framed, what's missing, and how it compares with other coverage of the same story. I couldn't find one that did all of that in one place, so I built it.
Lenstrum doesn't tell you what to think. It gives people the same set of facts to work from, so the conversation can keep going.
I'm building this on my own, one decision at a time. If any of this sounds like something you care about, I'd like to hear from you.
Founder, Vancouver, BC
I publish the methodology openly and want people to challenge it. If you spot a systematic issue with the analysis, use the feedback tools — they come straight to me.
© 2026 Lenstrum Analytics Inc. — Media Transparency Platform