Gibiru

By Bill Hartzer · Updated 19 August 2026

What Gibiru is, and where its results actually come from

Gibiru is a small American search site that describes itself as the only fully anonymous, uncensored search engine. What a user gets is a stripped-down results page with no cookies, no login and a toggle between ordinary results and an “uncensored” mode that is said to surface pages mainstream engines suppress. The pitch is explicitly political: the site has long asserted that government and platform pressure removes alternative-news outlets from mainstream results, and its audience comes from that framing rather than from technical privacy circles.

The verdict at the top of this page is resold results, and the reasoning is worth spelling out because Gibiru's marketing says the opposite.

An engine that runs its own crawler — the automated program that fetches pages and stores them in an index — has to tell the web that it exists. Site owners need a name to allow or block in robots.txt, hosting providers need a user-agent string to whitelist, and every independent engine of any size publishes a bot information page for exactly that reason. Gibiru has published none of it. There is no crawler name, no user-agent, no crawl-scope statement, no IP range and no index-size claim anywhere in seventeen years of operation. An engine that genuinely crawled the web for that long without ever naming its robot would be without precedent.

What the record does contain is a series of descriptions pointing the other way. In a 2012 interview the founder said Gibiru modifies Google's search algorithm so that the system does not use a visitor's IP address or cookie data when it returns results. In 2013 he described the product as built on open-source technology with a proxy search feature letting users pick exit IPs in different countries. Proxying is not crawling. In March 2022 an independent survey of search alternatives reported that all Gibiru results are actually routed through Google.

Where the sources disagree

This page prints both sides rather than averaging them: a genuine conflict in the record is more useful than a confident summary that hides it.

The own-index position. Cloudwards' guide to search engines that do not track states that Gibiru maintains its own index, including content not found in the curated feeds of Google and other popular engines, and presents results in two buckets — normal results and censored content. That is a current review from a mainstream consumer-technology publisher.

The Google-routed position. The founder's own 2012 and 2013 interviews describe a modified Google feed and a proxy, not a crawl. Larry Sanger's March 2022 survey began by finding no information about the sources of Gibiru's index and ended by concluding that all its results are routed through Google, with a recommendation to avoid it.

Reading the two together, the founder's own descriptions and an independent tester agree with each other, and the dissenting claim appears in a review that also concedes it could not identify who is behind the company. A second result set that filters or re-ranks an upstream feed is not an index; it is a different view of somebody else's index. On that basis the verdict stays at resold results, with the disagreement recorded rather than buried.

The “uncensored” mode

The uncensored toggle is Gibiru's headline feature and the reason most of its users are there. It is also the part of the product with the least documentation attached to it. There is no published methodology for what the uncensored result set includes, what it excludes, how it is compiled, or who decides.

Sanger tested the feature directly in 2022 and concluded that the label was strictly marketing with little substance behind it. He was fair about the exception: on one politically loaded test query, Gibiru returned more polling data, sociological and legal fact and fewer opinion pieces than several rivals. That is a real point in its favour and it is recorded here as such.

The structural criticism survives it. An undocumented result set assembled by undisclosed criteria, by an operator who has never published a methodology, is a worse epistemic position than a documented mainstream one, not a better one. A reader who distrusts Google's ranking because they cannot see inside it has not solved that problem by moving to a ranking nobody has described at all. “Uncensored” here means differently filtered.

Who runs it, and how it makes money

Gibiru was founded and is run by Steve Marshall, a resident of San Benito County, California and a former Navy special-operations veteran. In 2013 BenitoLink reported that it was operated from a virtual office by Marshall and partners, and run as a grassroots movement without a formal corporate structure. Mobile apps have been published under the developer name Steven Marshall.

Beyond that, corporate transparency is close to absent. No current legal entity, corporate registration or change of ownership could be located from any filing. A 2026-current review by Cloudwards makes the same observation, noting that it could not find much about the team behind the search engine. There is no team page, no named company, no transparency report and no warrant canary.

The jurisdiction point is under-reported and matters. Gibiru is a United States operation, and a US-domiciled service is reachable by US legal process, National Security Letters and the CLOUD Act. An engine whose entire marketing premise is escaping American government influence over search is itself domiciled in America.

On revenue, the current position is undocumented. Historically the model was advertising on results pages plus promotion of affiliated privacy products, with a VPN and browser and app downloads marketed alongside search. In December 2013 the founder told BenitoLink the company had not yet achieved profitability. No revenue figures, funding rounds, subscription tiers or API products are documented from a primary source. If the results are Google-derived, a search-ad revenue share plus affiliate income would be the natural model — but no statement confirming an advertising partnership could be sourced, so that is inference and is flagged as inference.

Privacy claims, and what can be verified

Gibiru's claims are the standard set: no logs, no tracking cookies, IP addresses not retained, 256-bit encrypted connections, anonymous by default. Third-party reviews restate them.

Verified findings: none. There is no independent audit, no transparency report, no warrant canary and no published security assessment of Gibiru. Every privacy claim on the site is self-attested. Unlike Startpage or Swisscows, there is not even a detailed policy document that can be read and archived without JavaScript, because gibiru.com serves a JavaScript application and renders nothing useful to a plain fetch. For a product whose entire value proposition is trust, having a privacy policy that is difficult to read and difficult to archive is a poor look.

Two structural points a reader deserves. First, if results are proxied from Google, then Google receives the query. Gibiru's protection is that Google does not receive your identity alongside it — the same de-identifying proxy model Startpage uses, which is a genuine protection but is not “Google never sees this.” Second, self-attestation from an operator whose legal entity cannot be identified is the weakest form of privacy assurance available. That is not an accusation of bad faith. It is a statement about what a user is being asked to accept on faith.

History and usage

2009 — Gibiru founded by Steve Marshall in San Benito County, California.
February 2012 — around 10,000 daily queries, company-reported.
7 May 2012 — RT covers Gibiru with the founder's anti-tracking framing and the site's claim that the NSA has had major publishers remove alternative-news sites from search results.
December 2013 — post-Snowden growth; over half a million monthly queries reported, with roughly 8 to 9 per cent month-over-month growth; still unprofitable and with no formal corporate structure.
c. 2019 onward — mobile apps published; the brand becomes a fixture of uncensored-search listicles and alternative-media recommendation lists.
March 2022 — Larry Sanger's widely read survey of search alternatives investigates Gibiru and advises against it on the grounds that results route through Google. This remains the most consequential public assessment of the engine (larrysanger.org).
2024–2026 — no company announcements, changelog or product news located. The site remains live as of 19 August 2026.

Both usage figures above are founder-supplied, unaudited and now thirteen to fourteen years old. They are historical colour, not a current estimate. There is no published current market share: Gibiru is not broken out in StatCounter's global or US search-engine data and sits inside the residual “Other” bucket. No lawsuits, breaches, acquisitions or regulatory actions were found, which is an absence of findings rather than a certification that none exist.

Who it suits, and who it does not

What it does well. It is a lightweight, no-account, no-cookie way to get mainstream-quality results without a mainstream login. Sanger's testing found relevance sometimes comparable to Brave and Qwant. It is fast, and it has run for seventeen years without degrading into an ad farm.

What it is bad at is the thing it most loudly claims. It is not a documented independent engine: there is no crawler, no index documentation, and the founder himself described a modified Google feed. “Uncensored” has no published methodology and could not be substantiated on independent testing. Corporate transparency is close to nil.

Who should look elsewhere. Anyone who believes they are escaping Google, because on the weight of the evidence they are not. Anyone with a real threat model — journalists protecting sources, activists, people at genuine risk — because unverifiable self-attestation from an operator with no identifiable legal entity is the wrong foundation for that. And anyone treating the uncensored toggle as an information-quality filter.

Stated plainly: on the available evidence Gibiru is a Google proxy with a political marketing layer. A reader who wants a Google proxy can have one from a company with a name, a published policy and a European data-protection regulator behind it. The functional case for preferring Gibiru is hard to locate.

Common misconceptions

“Gibiru has its own index of censored content”

The single most repeated claim about it, and it does appear in current reviews. No crawler, user-agent or index has ever been documented in seventeen years. The uncensored tab is an undocumented result set, not a demonstrated independent index.

“Gibiru is outside US jurisdiction”

It is a California operation, squarely inside the jurisdiction its own marketing warns about.

“It's anonymous, so Google never sees my search”

If the results are Google-derived, Google receives the query without your identity attached. De-identification, not invisibility.

“There's a Wikipedia article about it”

There is not. Gibiru appears on Wikipedia only in talk-page discussion for List of search engines. Third-party wiki mirrors and brand-profile pages are frequently mistaken for the real thing.

“Uncensored means unfiltered”

It means differently filtered, by undisclosed criteria, by an operator who has never published a methodology (Cloudwards and RT, 2012 are the two sources that disagree most sharply about what lies behind it).

Frequently asked questions

Does Gibiru have its own search index?

No documented one. In seventeen years Gibiru has never published a crawler name, a user-agent string, a crawl-scope statement or an index size, which independent engines routinely do so that site owners can identify their traffic. Its founder described the product in 2012 as a modified Google feed and in 2013 as a proxy. One 2026 review does claim an own index; that claim is disputed by the founder's own descriptions.

Is Gibiru just Google?

The weight of the evidence says the results are Google-derived, with tracking stripped out before they reach you. The founder said so in 2012, and Larry Sanger's 2022 independent survey concluded that all Gibiru results are routed through Google. Cloudwards disagrees and describes an own index. Gibiru itself has published no technical documentation that would settle the question either way.

Who owns Gibiru?

Steve Marshall, of San Benito County, California, founded it in 2009 and has been its public face since. Beyond that the picture is thin: in 2013 it was reported to run from a virtual office with no formal corporate structure, and reviewers as recent as 2026 report being unable to identify the team. No current legal entity or corporate registration could be located from any filing.

Is Gibiru safe and private to use?

Its claims — no logs, no tracking cookies, no retained IP addresses, encrypted connections — are entirely self-attested. There is no independent audit, no transparency report, no warrant canary and no published security assessment. The site is also fully JavaScript-gated, so even its policy text is hard to read or archive. Nothing suggests wrongdoing; there is simply no external verification of any kind.

What does Gibiru's uncensored mode actually do?

Nobody outside Gibiru knows. There is no published methodology for what the uncensored result set contains, what it leaves out, or how it is compiled. Larry Sanger tested it in 2022 and judged the label to be strictly marketing, while noting fairly that on one politically loaded query it surfaced more factual material and fewer opinion pieces than several competing engines.

Is Gibiru based outside the United States?

No. Gibiru is a California operation, which places it inside US jurisdiction and therefore within reach of US legal process, National Security Letters and the CLOUD Act. This is worth knowing because the site's marketing centres on American government interference in search results. The engine warning about that jurisdiction operates inside it.

Does Gibiru have a Wikipedia article?

No. There is no English Wikipedia article for Gibiru; it is mentioned only in archived talk-page discussion attached to the List of search engines article. Several third-party wiki-styled sites and brand-profile pages carry Gibiru entries and are regularly mistaken for Wikipedia. None of them is, and none should be cited as such.