From: jimruttshow8596
Zachary Vorhies, a former software engineer at Google who worked there for eight years, has made provocative claims against the company, alleging improper and illegal actions concerning search results, blacklists, and election meddling [00:00:20]. Vorhies publicly posted a large repository of what he describes as Google internal documents to support his claims [00:00:37].
Host's Caveat
The host, Jim Rut, stated he had not researched Zach Vorhies or reviewed his documents, nor did he have a strong opinion about Vorhies’ expertise and claims at the time of the interview [00:00:49]. He noted that while he strongly disagreed with some of Vorhies’ assertions, the claims about Google’s manipulation of the information sphere were disturbing if true [00:01:00]. The public can access the documents at projectveritas.com/google-dump [00:17:26].
Background and Motivation for Whistleblowing
Vorhies likened himself to “the Snowden of Google,” having collected documents for a few years because he believed Google was “lying to the American public” [00:01:39]. He claims Google was meddling in search results, YouTube queries, and elections [00:01:57]. He states he witnessed Google executives “perjuring themselves in front of congressional testimony” [00:02:07]. Vorhies specifically cited instances where executives told Congress they didn’t employ blacklists, yet he found internal documents and search results within Google for “blacklist,” including a “YouTube query blacklist” [00:02:20]. His decision to disclose was to inform the American people [00:02:42].
Following his disclosure, Google attempted to stop him by calling a “wellness check” on him [00:02:50]. Vorhies explained a “wellness check” as a tip to the police that someone is unwell or might do something, providing probable cause for police to search for and question the person [00:03:16]. He believes Google used a “Deadman switch” he posted on Twitter as justification for the wellness check [00:04:45].
Allegations of Political Bias and Election Meddling
Vorhies claims that after the 2016 US presidential election, Google executives “freaked out” [00:06:25]. During a weekly company meeting (TGIF), executives allegedly attributed Donald Trump’s election to Russia hacking the election and widespread racism, hate, and misogyny in the public [00:07:01].
Around December 2016, Vorhies noticed “odd papers” appearing on Google’s internal networks [00:07:51]. One such paper was titled “Psychological Induced Stress as a Result of Algorithmic Discrimination” [00:08:08]. He alleges this paper discussed how to alleviate user stress by providing “non-functional buttons and dials” that would “pretend to turn some of the censorship off,” thus restoring a user’s sense of control [00:08:31]. The paper also allegedly explored simulating “crowding” to increase product purchases via ads [00:09:01].
Vorhies claims the company then ramped up its efforts to define and combat “fake news,” stating executives explained it was a reason Trump was elected [00:10:15]. He alleges that internal papers listed examples of “fake news” that were, in fact, factual events, such as the claim about Hillary Clinton running “weapons soup and gauzy to fund Isis” [00:10:48]. He questioned why Google was involved in such politics and concluded they were “sculpting the information landscape” [00:11:15].
Machine Learning Fairness
Vorhies identified a new system Google was building called “machine learning fairness” (ML fairness) [00:12:15]. He describes this as an Orwellian term [00:12:22]. He alleges that ML fairness revises search results, biases YouTube results, and filters news, with Google Search, YouTube, and News being its primary targets [00:12:48]. The system allegedly takes training data as input and generates rules to rank content based on “fairness” [00:13:05].
A document on algorithmic unfairness, according to Vorhies, stated that if a search for “CEOs” returned mostly male results, “even if this represented objective reality,” it could still be classified as “algorithmically unfair” and warrant “product intervention” [00:13:25]. He interprets this as Google using “fairness” to “distort reality” and “change the nature of reality to make a better person” [00:14:02].
Furthermore, Vorhies claims that Google internal documents describe users as being “programmed” by the content they interact with [00:14:27]. He alleges Google believes it has a “moral obligation” to “program people to be better” in ways Google defines [00:16:34]. He expressed concern that Google, due to its pervasive presence (Chrome browser, Gmail, Android phones), has the scale to sculpt the information landscape and potentially “program” Americans, making escape difficult if they “turn evil” [00:15:36].
Examples of Alleged Interference
Vorhies provided several examples of alleged manipulation:
- Search Autocomplete Bias: He claims that typing “Trump body-count” into Google search will auto-predict the full phrase, despite it being “ridiculous” and having low search volume according to Google Trends [00:26:24]. In contrast, typing “Hillary body” will not auto-predict “Hillary body count,” despite her having a “horrendous body count list” and higher search volume for that term, as indicated by Google Trends data [00:26:36]. He asserts this indicates a political orientation in Google’s autocomplete [00:28:09].
- Ireland Election Interference: Vorhies claims Google interfered in Ireland’s election by banning the phrase “eighth amendment to the constitution of Ireland,” which is disclosed in the YouTube query blacklist [00:28:33].
- Brazil Election Interference: He states that the Supreme Court of Brazil ruled that Google interfered in their election, citing a contract between Google and an opponent of the current Brazilian president [00:28:51].
- Tulsi Gabbard Ad Account Suspension: During the Democratic debates, Tulsi Gabbard’s ad account was suspended for “suspicious activity” when people searched for her [00:39:07]. Vorhies links this to a security vulnerability that Google has allegedly refused to patch, which a group of insiders know about and exploit [00:40:01]. He describes a method where accounts similar to a target’s email (e.g.,
xxx@gmail.comtoxxy@gmail.com) are created and used to push spam, causing Google’s AI to deactivate the spam network and, “surprisingly,” the original target email address [00:40:36]. He claims to have seen a bug report confirming this behavior related to Jordan B. Peterson’s email account [00:41:12]. Vorhies asserts this is a pattern insiders use to take down targeted accounts [00:42:29]. - Trump Tweet Translation Manipulation: Vorhies’ “breaking point” was when he observed Google allegedly deleting translation words from the Arabic-to-English dictionary to make a Trump tweet sound “crazy” [00:43:22]. This perceived “craziness” was then allegedly used by The New York Times to try to invoke the 25th Amendment to remove the President [00:43:40]. The word in question, “calf F a,” which translates to “we will stand up,” was allegedly deleted twice because Google’s AI was “smart enough” to find a phonetic transliteration path [00:43:54]. Vorhies considers this “treasonous behavior” and a “national security issue” [00:45:46].
Claims of Perjury
Vorhies directly claims that Google executives have committed perjury to Congress [00:24:58]. He specifically cites a Google representative, Karen Buta, who testified under oath that Google does not employ blacklists [00:25:20]. Vorhies asserts that he has disclosed one blacklist, and Robert Epstein has disclosed nine previous ones, bringing the total to about twelve public blacklists [00:25:36]. He also claims the CEO made similar false statements under oath regarding political censorship [00:26:09].
Google’s Internal Culture and Corporate Structure
Vorhies states that Google’s culture began to decline in 2016, specifically after the election of Trump [00:05:58]. He observed that while engineers at Google are “so smart and so competent,” management appears to be “selected more for loyalty than for competence” [00:05:28]. He views this as “rot” setting in when managers “kiss up” rather than “punch up” at their bosses and protect their people [00:05:41].
He also discussed Google’s unusual corporate structure, noting that since its IPO, it has maintained different classes of stock, with Class B shares having disproportionate voting rights (e.g., ten to one) [00:22:41]. Vorhies asserts that Class B shareholders, who are insiders, “absolutely control the company” [00:24:07]. This structure means stockholders cannot use their voting power to prevent Google from allegedly abandoning its fiduciary responsibilities or engaging in political meddling [00:24:25]. He believes Google’s “flowery language” about its “high MIT morals stance” is a “cover for criminal activity” [00:24:42].
Whistleblower Information and Advice
Vorhies’ documents are available at projectveritas.com/google-dump (specifically the “everything.zip” file) [00:17:26]. He has also published articles detailing his claims, such as “Why did Google suppress autocomplete for Clinton body count” on Medium (also found on Y Combinator) [00:34:35]. He encourages people to test Google’s alleged bias by cross-referencing Google’s autocomplete suggestions with traffic data from trends.google.com [00:37:31].
For other Google employees or individuals with evidence of misdeeds, Vorhies advises against direct communication with him, as he believes he is being watched [00:46:51]. Instead, he recommends contacting Project Veritas, which he describes as a “professional media organization” capable of maintaining whistleblower discretion and protecting identity [00:47:06]. He suggests submitting information via their website’s “submit links” feature, preferably using a VPN and a Tor browser for security [00:48:25].