About Lens

Most of us have felt it. You read a headline and something feels off. You search for the same story on another site and it reads completely differently. You can't tell if what you're reading is fact, framing, or somewhere in between — and no one is telling you.

That feeling is what Lens was built to address.

We started with a simple problem. Finding a news source that was genuinely unbiased felt impossible. And when that search failed, a different question emerged — what if instead of hiding bias, we just showed it to you? What if we gave you the tools to read the news knowing exactly what angle each outlet is pushing, how many independent sources have confirmed a story, and what the facts actually are beneath the framing?

That is what Lens does.

We believe the news isn't broken. Framing is inevitable — every editorial decision, every word choice, every headline reflects a perspective. That isn't dishonest. It's human.

What is broken is the pretence that it isn't happening.

Lens doesn't claim to be unbiased. No source can make that claim honestly. Instead we give you something more useful — transparency. We show you who is reporting a story, what angle they're taking, and how confident you should actually be in what you're reading.

We believe informed people make better decisions. And we believe you deserve to be informed — not managed, not nudged, not kept scrolling.

Every story on Lens is built from multiple sources. When a story breaks, we track it across every outlet covering it and we show you three things no other news app shows you together in real time:

How reliable it is. A story reported by one outlet is unconfirmed. Reported by twelve independent sources including wire services, it's confirmed. Our reliability score updates automatically as coverage grows, so you always know how much weight to give what you're reading.

Who is saying it and from what angle. Every source covering a story carries a bias rating — from Left to Right, with Lean Left and Lean Right as intermediate positions. State-controlled and questionable sources are flagged separately. Not as a judgement, but as context. The facts of a story and the framing of a story are two different things. Lens shows you both.

Just the facts. Our AI reads every source covering a story and produces a single neutral summary — no opinion, no emotional language, no framing. A starting point, not a conclusion. Read it, then read the sources themselves knowing exactly what lens each one is looking through.

We are not a fact-checker. We are not arbiters of truth. We do not decide what is real and what isn't.

We are not an engagement platform. There are no algorithms designed to keep you reading, no outrage loops, no content chosen because it provokes a reaction. Lens is quiet by design.

We do not use images. Photography is an editorial choice — a picture of a politician or a conflict scene frames the story before you've read a word. On Lens, the words speak for themselves.

Every story on Lens earns its reliability score from three observable facts about its coverage — never from our opinion of the outlets involved.

• Independent reporting. We count how many sources have independently reported a story. When several outlets republish the same wire copy, that counts once — wide distribution is not the same thing as confirmation. You will still see every source listed; syndicated copies simply don't inflate the score.
• Verified origination. Stories originated or confirmed by anchor sources — wire services and public-service broadcasters whose business is primary reporting — score higher. These organisations put reporters in the room; their involvement is a strong signal the underlying facts have been witnessed, not relayed.
• Confirmation across the spectrum. When outlets from the left, right and centre all report the same core facts, that agreement is itself evidence. A story confirmed across the political spectrum scores higher than one confirmed only by outlets that lean the same way.

What the labels mean. Single source means exactly that — one independent report so far, which is normal for breaking news and not a mark against the story. Limited means corroboration is still thin. Medium and High mean a story has been independently confirmed, increasingly broadly. Scores update in real time as coverage develops.

What the score is not. It is not a verdict on whether a story is true, and it is not a rating of any outlet's honesty. It measures how well a story is corroborated right now — transparently, from facts you can check yourself by opening the source list.

Every news source has a perspective. Lens doesn't hide that — it shows it to you. Each source is given a bias rating that places it on a spectrum from Left to Right, or flags it as a special category where a left-right label would mislead.

The spectrum positions are: Left, Lean Left, Centre, Lean Right, Right. Some sources carry a separate flag instead of a spectrum position: State-controlled media (state-owned or state-directed outlets, such as TASS) and Questionable reliability (sources with documented accuracy or methodology concerns). Where no credible rating exists, a source is shown as Not yet rated — we do not guess.

Ratings are not made by Lens. They are derived from a combination of established, publicly documented bias raters — primarily AllSides, Ad Fontes Media, and Media Bias/Fact Check for international and US sources, and YouGov perception surveys and academic election-coverage studies for UK outlets. Where multiple raters agree, confidence is high. Where they diverge, confidence is lower and displayed as such. Each rating links to the raters it draws from, so you can verify it yourself.

Bias ratings describe editorial positioning — the consistent angle a publication takes in its coverage — not factual accuracy. A right-leaning source may report facts accurately. A centrist source may occasionally get things wrong. The two are separate, and Lens separates them: the reliability score measures corroboration across sources; the bias rating describes perspective.

Ratings are reviewed periodically and may not reflect recent editorial changes. State-controlled sources are excluded from cross-spectrum reliability calculations. Sources: AllSides (allsides.com/media-bias/ratings) · Ad Fontes Media (adfontesmedia.com) · MBFC (mediabiasfactcheck.com) · YouGov UK media perception surveys.

• Story clustering — When a new article arrives, AI decides whether it belongs to an existing story or starts a new one. Strict matching — only groups articles about the exact same specific event.
• Headline generation — When a new story is created, AI writes a neutral, factual headline — maximum 12 words, no emotional language.
• Summary generation — Once a story has multiple sources, AI reads all of them and produces a structured summary: a lead (the main fact), key facts (bullet list), and context (one sentence of background). Only facts from the source articles are included — no outside knowledge or speculation.

AI summaries are clearly labelled in the app. Always read original sources.