Why Lenz

The “is this actually true?” question shows up multiple times a week — in a group chat, on a news headline, halfway through writing something you’re about to publish. Lenz is built for that exact moment: turn the question into a sourced answer you can check yourself, and share with confidence.

A receipt, not an opinion.

Why not just Google it or ask ChatGPT?

You can ask any chatbot whether something is true — and you’ll get a confident-sounding answer. The model draws on whatever it absorbed during training, with no obligation to check its own claims against real sources. When it doesn’t know, it guesses.

You can also Google it — and end up with five tabs open and your own opinion to form.

Lenz is built differently:

  • Source-first, not memory-first. Every claim is checked against independently retrieved, scored, and cited sources. The evidence drives the conclusion — not the model’s prior beliefs.
  • A panel, not a single voice. Multiple AI models from different providers evaluate each claim separately. Different training data, different blind spots — one model’s hallucination is another’s red flag.
  • Engineered rigour at every step. The process doesn’t just “ask” a model for its opinion. Each stage — framing, research, debate, adjudication, conclusion — follows structured prompts that enforce citations, detect bias, and penalise unsupported assertions. Systematic by design, not by luck.

Real moments when Lenz helps

Next time you’re not sure — don’t scroll past it, don’t paste it into ChatGPT and hope. Verify it. Lenz shows the sources, shows the work, and you decide what to make of the answer.

And it really is hope: in our research on 1,000 real-world fact-checks, the five top frontier LLMs disagree on 67% of them.