Reliable Sources for Wikipedia: The 2026 Guide (Forbes, Daily Mail & Contributor Posts)

A Wikipedia page lives or dies on its sources, and the part that catches most companies off guard is that "good sources" and "Wikipedia-reliable sources" are two different lists. You can be covered by a famous masthead and still get cited zero times in your own article — because the byline was a contributor, the piece was a reprinted announcement, or the outlet sits on a list the community has formally agreed to distrust.

This guide is the reference we keep coming back to when we assess a source base. It explains how Wikipedia actually rates sources in 2026 — the Perennial Sources list, the reliable/situational/deprecated tiers, and the specific traps (Forbes contributors, Daily Mail, churnalism) that turn a confident clip list into a thin one. We use the real policy names throughout, because anyone you hire should speak this language, and you should be able to tell when they don't.

How Wikipedia rates a source: the Perennial Sources list

Wikipedia's headline rule for sourcing is the reliable-sources guideline, shortcut WP:RS. It asks three things of every citation: does the source have a reputation for editorial oversight and fact-checking, is it independent of the subject, and is it appropriate for the specific claim it's attached to. That's the theory. In practice, editors don't relitigate the reputation of The New York Times every time someone cites it — nor argue from scratch about whether a pay-to-publish portal counts.

So the community built a shortcut: the Perennial Sources list (shortcut WP:RSP, formally Wikipedia:Reliable sources/Perennial sources). It's a maintained table of the outlets that come up in disputes again and again, each with a consensus rating and a link to the discussions behind it. When an editor at a deletion debate sees a citation, the first thing they often do is check whether the outlet is on RSP and what colour it's been assigned. It is, effectively, a precedent database for sourcing.

A few things about how RSP works are widely misread:

• It's consensus, not a rulebook. Each entry summarises where community discussions have landed, not a top-down decree. Ratings change as outlets change — editorial collapses, ownership shifts, a reliable outlet spins up an unreliable contributor arm.
• It's not exhaustive. Most outlets aren't on RSP at all. Absence doesn't mean "unreliable" — it means "no recurring dispute," and the source is judged on the general WP:RS criteria instead.
• Reliability is contextual. An outlet can be reliable for one kind of claim and not another. A respected science journal is a poor source for celebrity gossip; a fashion magazine is a poor source for a regulatory finding.
• It's broadly shared across editions. English Wikipedia maintains the canonical list; other language editions have their own conventions, but the big international ratings (wire services, deprecated tabloids) travel widely.

The practical upshot: before anyone drafts your page, every prospective citation should be checked against RSP. Green is a strong signal; yellow means "expect an argument"; red or deprecated means "don't build anything load-bearing on it."

The three tiers: reliable, situational, deprecated

RSP sorts sources into bands. The exact wording has shaded over the years, but functionally there are three you need to recognise, plus an outright-blacklisted category for the worst offenders.

Two clarifications that save a lot of confusion. First, "generally reliable" is not a blank cheque — even a green outlet fails to support notability if the specific piece is a reprinted announcement or a sponsored insert (more on that below). The tier rates the outlet's newsroom, not every URL on its domain. Second, "deprecated" is stronger than "unreliable." An unreliable source is simply weak; a deprecated source has been through a formal community deprecation discussion (an RfC) and is treated as presumptively unusable, often with an edit filter that warns anyone trying to add it. Building a page on deprecated sources isn't just risky — it's a guarantee of removal.

The Forbes problem: staff articles vs. the contributor network

No outlet causes more confusion in our intake calls than Forbes — so it deserves its own section.

Here's the issue. Forbes runs two parallel content streams under one logo. There's Forbes staff journalism: reported, edited, fact-checked work by employed journalists, which the community generally treats as reliable. And there's the Forbes contributor network (and its imitators — Inc. contributors, Entrepreneur's contributor program, Fast Company contributor pieces, countless others): outside writers who self-publish columns through a platform with minimal editorial gatekeeping. On RSP, Forbes contributors are rated generally unreliable and explicitly treated as self-published sources.

That distinction matters enormously, because the contributor network is where a lot of "as featured in Forbes" coverage actually comes from. Founders write their own columns; PR agencies place client-friendly pieces under a contributor byline; advisors return the favour for a portfolio company. The URL says forbes.com; the substance is a self-published op-ed wearing a trusted logo. Wikipedia editors know this on sight, and the first thing they do with a "Forbes" citation is check the byline tag — "Contributor" vs. staff — and discount accordingly.

How to tell them apart yourself, before you ever send a source list:

• Look for the "Contributor" label under or beside the author's name. Staff writers are usually marked "Forbes Staff" or "Senior Writer."
• Check the disclaimer. Contributor pieces carry a standard line noting that opinions are the writer's own.
• Read the framing. Staff pieces report and attribute; contributor pieces often read like the subject's own thought-leadership, because frequently they are.

None of this means a genuine Forbes-staff feature is worthless — a reported, independent staff profile is a legitimate, often strong source. It means you can't count "Forbes" as a monolith. Every Forbes link gets sorted into one of two piles, and only one carries weight. If your strongest evidence is a stack of contributor columns, the honest read is that you have very little Wikipedia-usable Forbes coverage at all.

Daily Mail and the other deprecated outlets

The most famous cautionary case in Wikipedia sourcing is the Daily Mail. In 2017, after a lengthy community discussion, English Wikipedia editors voted to deprecate the Daily Mail as a source — the first time a mainstream, high-circulation newspaper had been formally ruled generally unreliable for factual claims, citing a poor record of fact-checking, fabrication, and sensationalism. The decision was reaffirmed in later discussions. Today, attempts to add a dailymail.co.uk citation trigger a warning filter, and existing ones get stripped or flagged.

The Daily Mail set a template, and the deprecated list has grown. It now includes, among others:

• Tabloids with poor fact-checking records — The Sun, the Daily Mirror (situational-to-poor), the National Enquirer.
• State-controlled or propaganda outlets — RT (formerly Russia Today), Sputnik, Press TV, Xinhua for politically sensitive claims, China's Global Times.
• Outlets with a track record of fabrication or extreme partisanship — Breitbart, The Epoch Times, The Gateway Pundit, InfoWars, The Daily Caller (situational, leaning poor).
• "Looks like a database" sources — LinkedIn, Crunchbase, and company-profile pages are treated as non-independent, user-generated, or directory listings, not reliable secondary coverage.

The lesson for a company isn't political; it's procedural. A citation can be removed not because the underlying fact is wrong, but because the source carrying it has been deprecated. If a chunk of your coverage sits in deprecated outlets — and it happens more than you'd think with tabloid lifestyle features or syndicated wire-pickups — that coverage does nothing for you, no matter how flattering. Worse, leaning on it signals to reviewers that the source base is thin, which invites a harder look at everything else.

Independence, secondary vs. primary, and "churnalism"

Reliability is only one axis. A perfectly reliable outlet can publish a piece that's useless for your page because it fails on independence or on being secondary. These two concepts do more quiet damage to source lists than outlet ratings ever do.

Independence asks who originated the content, not just who published it. A green-rated newspaper that prints a feature transparently rewritten from your press kit is not giving you independent coverage — the masthead is reliable, but the content came from you. Wikipedia's organizations guideline has a specific test for this, independence of content (shortcut WP:ORGIND): material substantially based on company-supplied information, interviews where the subject is the sole voice, and announcements relayed without independent reporting all fail it. Independence isn't about the logo; it's about whether a journalist did their own work.

Secondary vs. primary is the companion test. A primary source is one step from the raw event with no analysis: a press release, a regulatory filing, an earnings transcript, a verbatim Q&A. A secondary source is a journalist or analyst standing back to provide their own framing and assessment. Wikipedia is built almost entirely from secondary sources; primary ones can support a discrete fact ("founded in 2014") but cannot establish that you matter.

Which brings us to churnalism — journalism that's really just lightly-edited PR. A press release goes out; an outlet runs it with a fresh headline and a byline; it looks like reported coverage but contains no independent reporting. It is reliable-outlet-shaped and notability-worthless. The tells: it tracks your release almost paragraph for paragraph, the only quotes are from your people, there's no outside expert or contradicting view, and it appeared within hours of your release alongside near-identical versions elsewhere.

That last point connects to a rule that surprises everyone: one wire story reprinted across fifty sites is still one source. Syndicated coverage shares a single origin, so it collapses to a single data point no matter how many logos carry it. We've watched "30+ media mentions" shrink to two or three genuinely independent pieces once reprints and churnalism are stripped out — and editors at deletion review do that stripping fast.

Build a defensible source set BEFORE you draft

Here's the order-of-operations point that separates pages that survive from pages that get deleted: the source assessment comes before the draft, not after. Most expensive mistakes in this field come from writing first and sourcing second — the article gets built from whatever turns up on Google, and a third to a half of it gets cut at review when the citations don't hold.

The disciplined sequence we run, and the one any competent partner should run:

1. Inventory everything. Pull every meaningful piece of coverage from roughly the last 24–36 months into one list — URLs, outlet, byline, date.
2. Collapse the duplicates. Identify wire-service originals and syndication, and merge every reprint of the same underlying story into a single entry.
3. Rate each survivor on four axes. Is it reliable (check RSP)? Independent (who originated it — WP:ORGIND)? Secondary (analysis, not relay)? Significant (about you, in depth, not a passing mention)?
4. Sort into pass / borderline / fail. A clean pass clears all four. A borderline clears most and will be argued over. A fail doesn't count, and you shouldn't pretend it does.
5. Read the result honestly. Three or more independent, spread-over-time passes, and a page is realistic. One or two passes on a heap of borderlines is genuinely uncertain. All borderlines and fails means "not yet" — and the right move is to close the coverage gap first.

This is exactly the discipline behind a Wikipedia notability audit: we take your list, run each item through that grid, and hand back a written verdict with reasoning per source. The audit doubles as a defensible internal record — when leadership asks why a page didn't launch, you have one page showing precisely which sources fell short and what would close the gap. When the sources do support an article, Wikipedia page creation becomes a writing-and-review exercise rather than a gamble, because the hard question was answered before a word was drafted. And when the gap is real, the fix is an earned-media project — building the independent, secondary coverage that doesn't yet exist — not a Wikipedia project.

Sources that double as AI-trust signals

There's a payoff to all this rigour that goes well beyond the encyclopedia. The same source qualities Wikipedia demands — independent, reliable, secondary, sustained — are increasingly the qualities that determine whether AI systems trust and cite your brand.

Large language models and AI answer engines are trained on, and retrieve from, the open web, and they weight it unevenly. Wikipedia itself is one of the most heavily-weighted sources in modern training corpora, so a well-sourced page propagates into how ChatGPT, Gemini, Claude, and Perplexity describe you. But the effect runs deeper than the page: the underlying sources — a reported feature, an independent analysis, a regulatory record — are signals these systems read directly. A reliable-source footprint is, in effect, a trust footprint, read off the same web by editors and machines alike.

The inverse is just as true. A brand whose presence is dominated by press releases, contributor columns, and deprecated-tabloid pieces gives AI systems thin, low-trust, often contradictory material — which is exactly when models hedge, get details wrong, or decline to say much at all. The fix for "the AI describes us badly" is not a prompt trick or a vendor claiming to inject content into a model; nobody can do that. The fix is the same source infrastructure that earns a Wikipedia page. That's the throughline of our AI visibility work: build the independent, reliable, secondary record once, and both the encyclopedia and the answer engines improve together. A defensible source set isn't only about Wikipedia anymore — it's the substrate AI reads your reputation from.

A source-strength self-check

Run your strongest five to ten pieces of coverage through these questions. Be as honest as an experienced editor would be — it's cheaper to be honest with yourself now than to discover it at deletion review.

• Is the outlet rated generally reliable on RSP? If it's situational, expect an argument. If it's deprecated, strike it.
• If it's Forbes, Inc., or Entrepreneur — staff or contributor? Check the byline tag. Contributor pieces are self-published; they don't count.
• Did a journalist do independent work, or is this your press release in disguise? Look for outside framing, an external expert, a critical note. If the only quotes are yours, it's primary.
• Is it secondary? Analysis and reporting count. Releases, filings, transcripts, and straight Q&As are primary.
• Is it significant? The piece must be substantially about you — enough to write several encyclopedic sentences from — not a passing mention in an industry round-up.
• Is it genuinely independent of the syndication? If four of your "sources" are the same wire story on four sites, that's one source.
• Is the coverage sustained? Pieces spread across months or years beat a single spike around one announcement.

Tally it up. If five or more pieces clear every question — independent of each other, spread over time — you have a defensible base and a page worth pursuing. If most of your list dies on independence, contributor bylines, or syndication, you don't have a Wikipedia problem; you have a coverage problem, and the honest path is to build the record before commissioning the page. Either way, you now know what the reviewers will see when they open your sources — which is the whole point.

Want us to run your coverage list through this grid and tell you which sources actually hold up? Email team@wikibusines.com and we'll send back an honest, source-by-source read.