Google Leaks - Book Summary

What's at stake? Unveiling the secrets of a tech giant from an insider's perspective.

Can you imagine a world without Google? For the majority of us, it's quite a challenge. Whether we're searching for information, streaming videos on YouTube, or sending emails via Gmail – Google has seamlessly integrated into our daily lives.

But do we really comprehend the intricate layers of this corporation? For most, the answer is a resounding no. And that, according to Zach Vorhies, is a significant concern. As a former Google employee, he claims that the company has been operating behind-the-scenes to influence political direction since Donald Trump stepped into the presidency.

This narrative will follow Vorhies' journey from being an enthusiastic team member of Google to becoming a whistleblower, revealing alarming assertions against his once admired employer.

Throughout the course of this story, you’ll discover

how Google reacted to Donald Trump’s unexpected election victory,

why the innocuous word “covfefe” sent shockwaves through the Google workforce, and

what led to the surprising visit to the author's residence by a battalion of fully armed law enforcement officers.

Donald Trump's victory unveiled an unexpected side of Google.

It was a typical morning in November 2016 at Google's offices in San Bruno, California, until Zach Vorhies walked in. The previous night, Donald Trump had been elected President of the United States — a fact that seemed to instigate emotional turmoil among the employees.

Grief-stricken faces, tearful conversations that resembled the aftermath of a personal tragedy, and absence letters filled the office. Vorhies was astounded by the extreme reaction. In a democratic society, elections are a natural occurrence, a contest where your side might not always emerge as the winner. Accepting defeat graciously and striving for improvement should be the course of action, but this wasn't what he observed within Google's walls.

This pivotal moment: Donald Trump's election seemed to bring about a political paradigm shift within Google.

Wherever Vorhies turned, he heard colleagues grumbling about the perceived injustice of the election results and insisting on the need for resistance against Trump's administration. Initially, he dismissed this as mere talk. But everything changed when employees were asked to attend a company-wide meeting live-streamed from Google headquarters in Mountain View, California.

The gathering had an eerie tone. Google's workforce, many donning the company's iconic striped beanie with a playful propeller on top, suddenly appeared to Vorhies as a uniformed army of political activists.

During the meeting, Sergey Brin, Google's co-founder and CEO of its parent company, Alphabet, branded Trump's election as "deeply offensive." Kent Walker, Vice President for Global Affairs, associated the victory with xenophobia and animosity. Meanwhile, Chief Financial Officer Ruth Porat invited colleagues to comfort each other with embraces.

The most concerning moment came when a staff member queried how Google could combat the supposed misinformation and fake news that catalyzed Trump's victory. Google's CEO, Sundar Pichai, responded in a way that both perplexed and alarmed Vorhies. Pichai proposed leveraging this situation as an opportunity to enhance machine learning and artificial intelligence. These systems had been previously used to fight issues like bullying and toxic comments. However, once they were expanded, AI and machine learning would be utilized for more significant purposes.

This revelation marked Vorhies' first hint of Google's ominous intentions.

Under the guise of combatting fake news, Google allegedly facilitated covert censorship.

Growing up, Vorhies was a firm believer in free speech, considering it the lifeblood of a genuine democracy. His perspective on various matters was shaped by absorbing information from diverse sources and evaluating differing viewpoints. Occasionally, his beliefs were transformed by newfound knowledge or a compelling argument. This was, he believed, the essence of free speech.

However, Vorhies felt that Google was challenging this natural flow of information. He sensed a resistance within the company towards fostering open discussions. It seemed to him that, within the Google ecosystem, there were predefined 'correct' opinions, and the organization subtly guided its users towards adopting these viewpoints, one search at a time.

In essence: Vorhies perceived censorship hiding beneath Google's endeavor to battle fake news.

From Vorhies's perspective, Google's fight against "fake news" appeared more like an attempt to control the kind of news and arguments the public could access. It seemed eerily similar to censorship.

At this point, Vorhies only had a vague intuition of Google's intent. So he started digging. Luckily, Google presented itself as an "open" company, making most of its internal documents accessible to all employees. Vorhies decided to conduct a simple search for "fake news," and the first document he encountered shed some light on the company's plans.

The document highlighted five instances of fake news, with four criticising Hillary Clinton, and the lone one supporting Trump. Some of the stories categorized as "fake news" weren't evidently so, according to Vorhies. For example, a headline stated, "Wikileaks CONFIRMS Hillary Sold Weapons to ISIS." The veracity of this assertion isn't crystal clear. However, according to Vorhies, several news reports did seem to suggest that Clinton's foreign policy decisions indirectly armed adversarial factions.

The selection of these fake news examples seemed biased to Vorhies. The news pieces were predominantly pro-Trump and anti-Hillary, which questioned their objective and apolitical nature.

Being an engineer, Vorhies knew that distinguishing between real news and fake news would require an automated process — it wouldn't be done manually. Thus, he found himself pondering over how this new system would function.

Google's intent to redesign the Internet for 'fairness' sparked concerns.

During the first few months of Trump's administration, Vorhies stumbled upon a document discussing Google's strategy for refining its search results. They had named it "machine learning fairness," an appealing term to most.

Machine learning involves feeding algorithms with real-world data, such as human decisions, enabling them to discern patterns and make decisions independently. 'Fairness' in this context referred to an algorithm that could learn from human behavior without adopting our inherent prejudices and biases.

While this might sound uncontroversial at first, it holds deeper implications.

Here's the takeaway: Google's plan was to recreate the Internet to make it 'fair.'

As he delved deeper into the document, Vorhies discovered more. Alongside machine learning fairness, a related concept — algorithmic unfairness — was introduced. This notion was something that machine learning fairness aimed to mitigate.

Let's understand this better with an example from the document: If a search for "CEOs" yielded more images of men than women, it would be deemed as algorithmic unfairness, considering it discriminatory towards women.

Vorhies felt a pang of unease. What if there genuinely were more male CEOs than female? Wouldn't mirroring the reality be fair, rather than 'algorithmically unfair'?

The document suggested that Google might not agree. Occasionally, true fairness could mean portraying things as they ought to be, not as they are. In other words, even an accurate representation could be algorithmically unfair. Showcasing mostly male CEOs, for instance, could perpetuate harmful stereotypes about gender and leadership.

Google seemed to believe that helping society evolve into a fairer and more equal entity could sometimes take precedence over reflecting reality.

What would happen to websites that didn't align with this worldview? Perhaps they could be relegated in the rankings or stripped of their capacity to earn ad revenues.

As per the documents Vorhies leaked, Google aimed to create a team of evaluators, working alongside AI, to assess the trustworthiness of various sites. However, would this process be impartial and balanced? Vorhies had his doubts. He suspected that Google might favor websites that echoed their viewpoints while downplaying those they disagreed with.

To Vorhies, this program seemed potent, capable of significantly altering American politics. The implications of such a shift deeply disturbed him.

Google's handling of the infamous 'covfefe' situation painted Trump in an unfavorable light.

In the early hours of May 31, 2017, a cryptic tweet from Donald Trump sparked worldwide curiosity. It read, "Despite the constant negative press covfefe." The perplexing term "covfefe" left the public scratching their heads, wondering what it could mean. Trump's press secretary, Sean Spicer, responded to inquiries with an enigmatic statement, claiming that the president and "a small group of people" knew precisely what it signified. But was this true?

Interestingly, Google's translation service seemed to unlock the mystery. Vorhies and others observed that Google Translate identified 'covfefe' as an Arabic phrase, which means "we will stand up." So, the tweet could read, "Despite the constant negative press, we will stand up." To Vorhies, it was a reasonable explanation. But Google wasn't pleased with this interpretation.

Here's the crux: Google's reaction to the 'covfefe' incident appeared to deliberately depict Trump in a negative light.

The first attempt to discredit Google's translation of 'covfefe' surfaced from the New York Times. Journalist Liam Slack, the following day, ridiculed the assumption that Trump had intentionally used an Arabic term.

Slack pointed out that Trump had previously vowed to prohibit Muslims from entering the United States. He also noted that Trump had never publicly claimed fluency in Arabic. Additionally, an Arabic professor clarified that 'covfefe' was entirely nonsensical.

This backlash was the cue Google had been waiting for. Vorhies pieced together the ensuing actions through the company's internal documents.

Google employees decided to alter the original translation of 'covfefe.' In an amusing move, they programmed Google Translate to display a shrugging man emoji when users input the word 'covfefe.'

With a few mouse clicks, Google employees stripped 'covfefe' of its Arabic translation, rendering Trump's tweet incomprehensible once more.

Although not the most serious or shocking act by Google, Vorhies found it highly indicative. The readiness and haste with which the company altered the translation was simultaneously impressive and worrisome.

The unearthing of 'blacklists' drove Vorhies to expose Google's activities.

The 'covfefe' incident was just the tip of the iceberg. What really ruffled Vorhies' feathers was his discovery of several 'blacklists.' These were, as he believed, lists of terms and websites that Google was deliberately striving to bury.

At that point, Google was proclaiming to the world that it never let political biases interfere with its search results. However, a simple search through Google's internal servers revealed documents with 'blacklist' in their titles — documents that appeared to target conservative media primarily.

Famous right-leaning websites like True Pundit, Louder with Crowder, and GlennBeck.com were featured on a list that, according to Vorhies, was intended to exclude content from Google Now newsfeeds. This revelation was his breaking point — he knew he had to bring this to the public's attention.

To be clear: the discovery of 'blacklists' urged Vorhies to expose Google's practices.

Vorhies reached out to the investigative news group Project Veritas — a right-wing organization known for its dramatic political exposés — after conducting some research. He started meeting with a representative from Project Veritas, meticulously explaining how he thought Google was systematically suppressing conservative content. Talking to the journalist was immensely cathartic, yet frustratingly, his claims were met with silence. No investigations or news reports were initiated. Not even a single tweet was made. For months, Project Veritas sat idle on the information.

During this time, Vorhies decided he could no longer be a part of Google. He felt that the company had betrayed its original motto — "Don't be evil." Disheartened by his job and the nonaction of Project Veritas, Vorhies chose to leave Google for good.

Then, abruptly, things took a turn. Project Veritas secretly recorded a Google executive stating that smaller digital organizations lacked the means to "prevent another Trump situation." Was that how Google viewed its role? To Vorhies, the recording affirmed his suspicions, and Project Veritas seemed to agree.

Project Veritas compiled a report on Google's alleged political bias, featuring an anonymous interview with Vorhies. Finally, his revelations were starting to yield results.

As Google retaliated against Vorhies, the situation heated up — however, at long last, the truth had emerged.

Vorhies had pulled it off. After years of harboring doubts, he'd finally been able to expose his findings to the world. It was a relief, no doubt — but paradoxically, this revelation sparked new anxieties.

Though he had maintained anonymity during his Project Veritas interview, Vorhies was aware that Google could easily uncover his identity. The company would have digital logs recording the identities of individuals who accessed and saved the internal documents he'd stumbled upon. It wouldn't take long for them to piece together the puzzle.

Therefore, it didn't surprise him when he received a cease-and-desist order from Google shortly after the exposé was published.

Here's the crucial point: As Google retaliated against Vorhies, the situation heated up — but, at last, the truth had emerged.

Google's notice didn't merely demand that Vorhies cease divulging confidential information. It also commanded him to return his company laptop, which housed all the documents supporting his contentious allegations.

But Vorhies was convinced that the world needed to see the documents he'd unearthed. So, instead of handing over his laptop as per Google's request, he sent it — along with 950 pages of internal documents — to the US Department of Justice. He also alerted a contact at Project Veritas, instructing him to release the documents publicly in the event of Vorhies's "untimely death" — an arrangement that he announced to the world via Twitter.

Paranoia, perhaps? However, Vorhies did experience something quite alarming shortly afterwards. He alleges that Google attempted intimidation tactics in August of 2019 by requesting San Francisco police to conduct a "wellness check" on their former employee.

When Vorhies refused to interact with the police, the situation spiraled. His apartment was surrounded by armed police officers, observed by a helicopter hovering above. There was even a bomb disposal robot at his front gate. Eventually, Vorhies complied, coming out of the incident unharmed.

Now that Google knew his identity, he saw no point in preserving the facade of anonymity. He decided to sit for another interview with Project Veritas, this time, revealing his true identity. Even more impressively, the organization published the documents he had acquired — all 950 pages of them.

About a week following the police incident, Project Veritas released the second interview. This time, it wasn't with an anonymous whistleblower but with Zach Vorhies — the ex-Google employee.

Closing thoughts

The central idea here is:

Zach Vorhies, during his tenure at Google, grew increasingly alarmed at the company's perceivable shift towards political bias and conscious censorship. Following Donald Trump's ascension to power, Vorhies observed the tech titan seemingly chart a politically motivated course involving "blacklists," AI, and the disregard of contrary views as "fake news." Ultimately, these concerns weighed heavily on him, driving him to disclose his claims publicly, unveiling Google's internal workings.