Three Years of Misery Inside Google, the Happiest Company in Tech

On a bright Monday in January 2017, at 2:30 in the afternoon, about a thousand Google employees—horrified, alarmed, and a little giddy—began pouring out of the company’s offices in Mountain View, California. They packed themselves into a cheerful courtyard outside the main campus café, a parklike area dotted with picnic tables and a shade structure that resembles a giant game of pickup sticks. Many of them held up handmade signs: “Proud Iranian-American Googler,” “Even Introverts Are Here,” and of course, “Don’t Be Evil!” written in the same kindergarten colors as the Google logo.

After a few rounds of call-and-response chanting and testimonials from individual staffers, someone adjusted the rally’s microphone for the next speaker’s tall, lanky frame. Sundar Pichai, Google’s soft-spoken CEO of 15 months, stood in the small clearing in the dense crowd that served as a makeshift stage. “Over the last 24 to 48 hours, we’ve all been working very hard,” he said, “and every step of the way I’ve felt the support of 60,000 people behind me.”

It was, to be precise, January 30; Donald Trump’s presidency was 10 days old. And Executive Order 13769—a federal travel ban on citizens from Iran, Iraq, Libya, Somalia, Sudan, Syria, and Yemen, and a wholesale suspension of US refugee admissions—had been in effect for 73 hours, trapping hundreds of travelers in limbo at the nation’s airports. For the moment, the company’s trademark admonition against evil was being directed at a clear, unmistakably external target: the White House.

To all the world it looked as if Google—one of the most powerful, pro-immigrant, and ostensibly progressive corporations in the United States—was taking a unified stand. But that appearance of unanimity masked a welter of executive-level indecision and anxiety. It probably would have been more apt if Pichai had said that, over the previous 48 hours, he had been backed into a corner by thousands of his employees.

In those first days of the Trump era, Google’s leaders were desperate to avoid confrontation with the new regime. The company’s history of close ties to the Obama administration left executives feeling especially vulnerable to the reactionary movement—incubated partly on Google’s own video platform, YouTube—that had memed, rallied, and voted Trump into office. (It didn’t help that Eric Schmidt, then executive chairman of Google’s parent company, Alphabet, had been an adviser to Hillary Clinton’s campaign, or that some 90 percent of political donations by Google employees had gone to Democrats in 2016.) Kent Walker, Google’s risk-averse vice president of public policy, had been advising staffers not to do anything that might upset Steve Bannon or Breitbart. So when the travel ban was announced on the afternoon of Friday, January 27, Google executives initially hoped to “just keep [their] heads down and allow it to blow over,” according to an employee who was close to those early calculations.

But the tribal dictates of Google’s own workforce made lying low pretty much impossible. Larry Page and Sergey Brin, the former Montessori kids who founded Google as Stanford grad students in the late ’90s, had designed their company’s famously open culture to facilitate free thinking. Employees were “obligated to dissent” if they saw something they disagreed with, and they were encouraged to “bring their whole selves” to work rather than check their politics and personal lives at the door. And the wild thing about Google was that so many employees complied. They weighed in on thousands of online mailing lists, including IndustryInfo, a mega forum with more than 30,000 members; Coffee Beans, a forum for discussing diversity; and Poly-Discuss, a list for polyamorous Googlers. They posted incessantly on an employee-only version of Google+ and on Memegen, an internal tool for creating and upvoting memes. On Thursdays, Google would host a company-wide meeting called TGIF, known for its no-holds-barred Q&As where employees could, and did, aggressively challenge executives.

All that oversharing and debate was made possible by another element of Google’s social contract. Like other corporations, Google enforces strict policies requiring employees to keep company business confidential. But for Google employees, nondisclosure wasn’t just a rule, it was a sacred bargain—one that earned them candor from leadership and a safe space to speak freely about their kinks, grievances, and disagreements on internal forums.

Finally, to a remarkable extent, Google’s workers really do take “Don’t Be Evil” to heart. C-suite meetings have been known to grind to a halt if someone asks, “Wait, is this evil?” To many employees, it’s axiomatic: Facebook is craven, Amazon is aggro, Apple is secretive, and Microsoft is staid, but Google genuinely wants to do good.

All of those precepts sent Google’s workforce into full tilt after the travel ban was announced. Memegen went flush with images bearing captions like “We stand with you” and “We are you.” Jewglers and HOLA, affinity groups for Jewish and Latinx employees, quickly pledged their support for Google’s Muslim group. According to The Wall Street Journal, members of one mailing list brainstormed whether there might be ways to “leverage” Google’s search results to surface ways of helping immigrants; some proposed that the company should intervene in searches for terms like “Islam,” “Muslim,” or “Iran” that were showing “Islamophobic, algorithmically biased results.” (Google says none of those ideas were taken up.) At around 2 pm that Saturday, an employee on a mailing list for Iranian Googlers floated the possibility of staging a walkout in Mountain View. “I wanted to check first whether anyone thinks this is a bad idea,” the employee wrote. Within 48 hours, a time had been locked down and an internal website set up.

Employees also spent the weekend protesting as private citizens, out in the open. At San Francisco International Airport, a handful of Google lawyers showed up to offer emergency representation to immigrants; many more staffers joined a demonstration outside the international terminal. But one Googler in particular made national newscasts. On Saturday night, without informing anyone at Google, Sergey Brin showed up at the airport to join the crowds. He offered no other comment to the press except to tell Forbes, “I’m here because I’m a refugee,” and to make clear that he was there in a personal capacity.

Between pressure from employees and Brin’s trip to the airport—which had effectively committed the company to sticking its neck out—Google’s own official calculations began to shift. Over the course of the weekend, the company matched $2 million in donations raised by employees for crisis funds for immigrants’ rights. And then on Monday, at the last minute, Pichai decided to speak at the employees’ demonstration.

In his short, off-the-cuff remarks to the packed courtyard, Pichai called immigration “core to the founding of this company.” He tried to inject a dose of moderation, stressing how important it was “to reach out and communicate to people from across the country.” But when he mentioned Brin’s appearance at the airport, his employees erupted in chants of “Ser-gey! Ser-gey! Ser-gey!” Brin finally extricated himself from the crowd and shuffled up to the mic, windbreaker in hand. He, too, echoed the protesters’ concerns but tried to bring the heat down. “We need to be smart,” he said, “and that means bringing in folks who have some different viewpoints.” As he spoke, a news chopper flew overhead.

And that was pretty much the last time Google’s executives and workers presented such a united front about anything.

As the Trump era wore on, Google continued to brace itself for all manner of external assaults, and not just from the right. The 2016 election and its aftermath set off a backlash against Silicon Valley that seemed to come from all sides. Lawmakers and the media were waking up to the extractive nature of Big Tech’s free services. And Google—the company that had casually introduced the internet to consumer surveillance, orderer of the world’s information, owner of eight products with more than a billion users each—knew that it would be an inevitable target.

But in many respects, Google’s most vexing threats during that period came from inside the company itself. Over the next two and a half years, the company would find itself in the same position over and over again: a nearly $800 billion planetary force seemingly powerless against groups of employees—on the left and the right alike—who could hold the company hostage to its own public image.

In a larger sense, Google found itself and its culture deeply maladapted to a new set of political, social, and business imperatives. To invent products like Gmail, Earth, and Translate, you need coddled geniuses free to let their minds run wild. But to lock down lucrative government contracts or expand into coveted foreign markets, as Google increasingly needed to do, you need to be able to issue orders and give clients what they want.

For this article, WIRED spoke with 47 current and former Google employees. Most of them requested anonymity. Together, they described a period of growing distrust and disillusionment inside Google that echoed the fury roaring outside the company’s walls. And in all that time, Google could never quite anticipate the right incoming collision. After the travel ban walkout, for example, the company’s leaders expected the worst—and that it would come from Washington. “I knew we were snowballing toward something,” a former executive says. “I thought it was going to be Trump calling us out in the press. I didn’t think it was gonna be some guy writing a memo.”

II.

In a lot of ways, Google’s internal social networks are like a microcosm of the internet itself. They have their filter bubbles, their trolls, their edgelords. And contrary to popular perception, those networks are not all populated by liberals. Just as the reactionary right was rising on YouTube, it was also finding ways to amplify itself inside Google’s rationalist culture of debate.

For some time, for instance, one of the moderators of the company’s Conservatives email list was a Chrome engineer named Kevin Cernekee. Over the years, Google employees have described Cernekee fairly consistently: as a shrewd far-right provocateur who made his presence felt across Google’s social network, trolling both liberals and conservatives.

In August 2015, the giant IndustryInfo mailing list broke into a roiling debate over why there were so few women in tech. The previous year, Google had become the first Silicon Valley giant to release data on the demographics of its workforce—and revealed that 82 percent of its technical workers were male. To many inside the IndustryInfo thread, the number constituted clear and galling evidence that Google had to change. When the conversation devolved into a brawl over the merits of diversity—one that Cernekee joined—a senior vice president at Google attempted to shut it down. Cernekee proceeded to bombard the executive’s Google+ page with posts about his right to critique the pro-diversity “Social Justice political agenda.” “Can we add a clear statement of banned opinions to the employee handbook,” he wrote, “so that everybody knows what the ground rules are?” In response, Google HR issued Cernekee a written warning for “disrespectful, disruptive, disorderly, and insubordinate” comments.

Google also took action against employees on the opposite side of the debate for their conduct in the same thread; but disciplining Cernekee had more lasting consequences. In November 2015, Cernekee filed a charge with the National Labor Relations Board claiming that Google’s warning constituted retaliation for his political views. He also alleged that the reprimand interfered with his right to engage in “protected concerted activity”—essentially, his right to freely discuss workplace conditions—as defined under the National Labor Relations Act.

As Cernekee entered into a years-long legal battle with Google, he stayed active on internal channels. In 2016, when members of a white nationalist group called the Golden Gate Skinheads clashed with antifa counterprotesters in a Sacramento park, Cernekee spoke up for the former on Google’s Free Speech mailing list. Though he said he was “the farthest thing possible from a Nazi,” Cernekee argued that the skinheads “stood up for free speech and free association.” And in January 2017, when the prominent white nationalist Richard Spencer was punched in the head by a masked protester after Trump’s inauguration, Cernekee told his fellow list members that “the battle over free speech is escalating.” He asked them to donate to a WeSearchr campaign that was raising a bounty for anyone who could track down the identity of the assailant. When mailing list members said that WeSearchr—a far-right answer to GoFundMe founded by the agitators Charles C. Johnson and Pax Dickinson—seemed shady, Cernekee wrote, “It is completely on the up-and-up. Please don’t slander my friends. :-(.”

But as conspicuous as Cernekee was inside Google, he was all but invisible on the open internet. Consequently, it wasn’t Cernekee who would become Google’s most famous heretic. That distinction would fall to a comparatively reticent Google Search engineer named James Damore.

In late June 2017, Damore attended a company event about promoting diversity at Google, hosted at the Mountain View headquarters. There, he claims, he heard organizers discuss providing extra job interviews and more welcoming environments for women and underrepresented minorities. (Google says it does not provide additional interviews for people belonging to specific demographics.) To Damore, this all sounded like a violation of Google’s meritocratic hiring process, a finely tuned system built to identify objectively qualified engineers.

Soon after, on the plane ride back from a work trip to China, Damore wrote a 10-page memo arguing that biological differences could help explain why there were fewer female engineers at Google, and therefore the company’s attempts to reach gender parity were misguided and discriminatory toward men. On average, he wrote, women are more interested in people than things, more empathetic, more neurotic, and less assertive. To support these claims about personality differences, Damore cited two studies, three Wikipedia pages, and an article from Quillette, a contrarian online magazine that often covers free speech on campus and alleged links between genetics and IQ. In the memo, Damore wrote that hiring practices aimed to increase diversity “can effectively lower the bar” at Google.

All through July, Damore tried to get Google’s management to pay attention to his concerns. He sent his memo to the diversity summit’s organizers; he sent it to Google’s human resources department; at the suggestion of a coworker, he posted it in Coffee Beans, the internal listserv for discussions about diversity. He made the same points in person at one of Google’s “Bias Busting” workshops, where employees role-play how to identify unconscious bias against minorities. (There, he later claimed, his coworkers laughed at him.)

Damore framed his memo as an appeal for intellectual diversity, identifying his reasoning as a conservative political position silenced by Google’s “ideological echo chamber.” “It’s a perspective that desperately needs to be told at Google,” Damore wrote.

Plenty of Damore’s colleagues, however, had heard this perspective before. Ad nauseam. “People would write stuff like that every month,” says one former Google executive. When the subject of diversifying Google’s workforce comes up in big meetings and internal forums, one black female employee says, “you pretty much need to wait about 10 seconds before someone jumps in and says we’re lowering the bar.” (After one diversity town hall in April 2015, an employee wrote in an internal Google+ post that Google was “lowering the hiring bar for minorities, or arranging events where white men feel excluded.”) What’s more, the debate kept coming up in a repetitive loop because of the constant influx of young graduates who were engaging in these discussions for the first time. Google was hiring at a breakneck pace at the time. Between 2015 and 2017, it added some 20,000 full-time employees, about the same number as Facebook’s entire workforce. (And even after all that hiring, Google’s technical workforce was 80 percent male, 56 percent white, and 41 percent Asian.)

Damore’s memo might have faded into obscurity if a colleague hadn’t suggested that he share it with some more receptive audiences inside Google. On Wednesday, August 2, Damore posted his memo to an internal mailing list called Skeptics. The next day he shared it with Liberty, an internal list for libertarians—one Damore hadn’t known existed. By Friday, the tech blog Motherboard was reporting that an “anti-diversity manifesto” had gone viral inside Google.

Pichai was on vacation when his deputies told him that Google had better deal with the Damore situation quickly. Pichai agreed and asked to corral his full management team for a meeting. By Saturday, a full copy of Damore’s document had leaked to Gizmodo. While Googlers waited for an official response from the top, managers who wanted to signal their support for women loudly condemned the memo’s ideas on internal Google+ posts.

To Liz Fong-Jones, a site reliability engineer at Google, the memo’s arguments were especially familiar. Google’s engineers are not unionized, but inside Google, Fong-Jones essentially performed the function of a union rep, translating employee concerns to managers on everything from product decisions to inclusion practices. She had acquired this informal role around the time the company released Google+ to the public in 2011; before launch, she warned executives against requiring people to use their real names on the platform, arguing that anonymity was important for vulnerable groups. When public uproar played out much as Fong-Jones had predicted, she sat across from executives to negotiate a new policy—then explained the necessary compromises to irate employees. After that, managers and employees started coming to her to mediate internal tensions of all sorts.

As part of this internal advocacy work, Fong-Jones had become attuned to the way discussions about diversity on internal forums were beset by men like Cernekee, Damore, and other coworkers who were “just asking questions.” To her mind, Google’s management had allowed these dynamics to fester for too long, and now it was time for executives to take a stand. In an internal Google+ post, she wrote that “the only way to deal with all the heads of the medusa is to no-platform all of them.”

A few hours later, Google’s internal networks received a shock to the system. A screenshot of Fong-Jones’ “Medusa” comment appeared on Vox Popoli, a blog run by the alt-right instigator Theodore Beale, along with her full name and profile photo. The comments section quickly filled with racial and sexual slurs fixated personally on Fong-Jones, who is trans. “They should pitch all those sexual freaks off of rooftops,” one anonymous Vox Popoli commenter wrote.

On Monday morning, Google’s top management finally met to discuss what to do about Damore. The room, according to reporting by Recode, was split. Half the executives believed Damore shouldn’t be fired. Then YouTube CEO Susan Wojcicki and head of communications Jessica Powell urged their colleagues to consider how they would have reacted if Damore had applied the same arguments to race, rather than gender. That persuaded them: The engineer had to go. In a note to employees, Pichai said he was firing Damore for perpetuating gender stereotypes.

In his message, Pichai tried to assure the left without alienating the right. “To suggest a group of our colleagues have traits that make them less biologically suited to that work is offensive and not OK,” he wrote. “At the same time, there are co-workers who are questioning whether they can safely express their views in the workplace (especially those with a minority viewpoint). They too feel under threat, and that is also not OK. People must feel free to express dissent.” Then he promised to fly back to the Bay Area for a TGIF meeting on Thursday where the employees could discuss the matter.

Damore’s termination set off a predictable onslaught of negative coverage from the right. Tucker Carlson, Ann Coulter, and Ben Shapiro all bashed Google; New York Times columnist David Brooks called for Pichai to resign. Damore gave his first interviews to YouTubers Jordan Peterson and Stefan Molyneux, the latter of whom is a proponent of “race realism.” The alt-right took this as an endorsement and started churning out memes of Damore, his head on Martin Luther’s body, nailing his memo to the door of a church.

More leaks from inside Google fed the frenzy. Screenshots of conversations among Google employees on internal social networks, some dating back to 2015, appeared on Breitbart. Meanwhile, on a pro-Trump subreddit, a collage appeared that showed the full names, profile pictures, and Twitter bios of eight Google employees, most of them queer, transgender, or people of color. Fong-Jones was one of them. Each bio featured phrases that would make the employees instant targets for online harassment: “polyamorous queer autistic trans lesbian,” “just another gay communist site reliability engineer,” or, in Fong-Jones’ case, “Trans and queer as fuck.” Two days after Damore was fired, Milo Yiannopoulos, the former tech editor at Breitbart, shared the Reddit collage image with 2 million Facebook followers. “Look at who works for Google, it all makes sense now,” he wrote—as if these eight employees had been the ones who made the decision to ax Damore.

For the employees who were being targeted, the leaks were terrifying. How many of their coworkers were feeding material to the alt-right? How many more leaks were coming? And what was their employer going to do to protect them?

In the past Google had fired an employee for leaking internal memes from Memegen. But when the targeted employees reported harassment, they say, Google’s security team told them that the leaking of screenshots might fall under the legal definition of “protected concerted activity”—the same labor right invoked by Cernekee.

To Fong-Jones, the security team’s answer was both shocking and instructive; she didn’t realize a leaker could be protected. “Everyone thought Google had an absolute right to stop you from talking about anything related to Google,” she says. Yet here Google’s hands were apparently tied by labor law.

Executives felt that they were doing everything they could. They offered to put up employees who had been doxed in a hotel for the night. But to the targeted employees, it felt as though Google was allowing fear of a far-right backlash and the threat of further lawsuits to trump the safety concerns of loyal employees. (A spokesperson for Google declined to specify whether the company managed to uncover the identity of the leakers in this case but said it investigates all such incidents and enforces its policies without regard to politics.)

By the day when Pichai was supposed to answer questions about Damore at TGIF, chaos surrounded Google. That afternoon, Damore returned to campus with a photographer whom The New York Times had recently dubbed “the Annie Leibovitz of the alt-right.” “Live in Mountain View,” the engineer tweeted, teasing some impending spectacle to his 40,000 new Twitter fans. Punchy tech reporters, sensing an opening, solicited employees to send them a live readout from the town hall.

Forty-five minutes before the meeting was supposed to start, Pichai sent his 78,000 employees an email. “TL;DR Sorry for the late notice, but we are going to cancel today’s Town Hall,” he wrote. “We had hoped to have a frank, open discussion today as we always do to bring us together and move forward.” But the questions employees had submitted for discussion at the meeting, he wrote, had been leaked to the press and were already appearing in outlets like WIRED. He also alluded, vaguely, to the employees who had been doxed as part of his rationale for canceling the meeting. “On some websites Googlers are now being named personally,” Pichai wrote.

Fong-Jones was at home in Brooklyn when she received Pichai’s email. She had wanted the executives to explain why, given their reasons for firing Damore, they had let his memo float around Google for more than a month. Now it felt as if Google was using the abuse she and other employees had experienced as an excuse to not answer questions. “That 100 percent did not sit right with me,” she says. “That wasn’t going to make me any safer. In fact, it was a triumph, almost, for the people harassing me.”

In her internal advocacy work, Fong-Jones had always been happy to meet senior managers on their terms. She kept their secrets. She followed the rules. And she instructed others to do the same. Executives talked to people they trusted, and they didn’t trust people who talked to outsiders. But now Fong-Jones decided to take matters into her own hands. In October she invited a labor group that normally organizes blue-collar workers to teach Googlers more about protected concerted activity. Perhaps a knowledge of labor law would come in handy.

III.

In the beginning, Google gave its employees freedom because doing so paid off. Some of the company’s first multibillion-dollar wins came from granting engineers autonomy and stoking their ambitions. In 2002 five engineers coding through the weekend came up with the core insight behind AdWords, the source of much of the company’s revenue. The engineers weren’t on the ads team and hadn’t been asked to work on the problem. But they saw a note Larry Page had left in the office kitchen on a Friday afternoon saying “THESE ADS SUCK.” By 5 am Monday the engineers had sent out a link with a prototype of their proposed solution.

That breakthrough set the stage for what seemed like a virtuous cycle. Because Google made its money by showing people ads, it had a vested interest in growing the internet itself, and in finding new ways to make it useful. Free, world-changing products poured out: Search, Gmail, Chrome, Maps, Docs, Photos, Translate. To the lucky beneficiaries—the company’s billions of users—it was almost as if Google were a public works department and not a capitalist enterprise at all. The best engineers on earth flocked to enlist, and the company offered them time to spend on their own larks. Google saw office perks, employee freedom, and lofty missions as a proven recipe for staying ahead in a changing world.

Over time, Google’s leaders codified this radical culture and evangelized it to the outside world, as if Google had found a way to suspend the ordinary laws of management and commerce. In their best-selling 2014 book, How Google Works, Schmidt and Jonathan Rosenberg, two of the main architects of Google’s culture, stressed the importance of open debate in the care and feeding of innovative people. “In our experience, most smart creatives have strong opinions and are itching to spout off; for them, the cultural obligation to dissent gives them the freedom to do just that,” they wrote. They stressed the importance of rooting out “knaves” (liars, cheaters, loafers) but supporting and protecting “divas,” difficult but brilliant employees who can grate on other employees’ nerves. “You need these aberrant geniuses because they’re the ones that drive, in most cases, the product excellence,” Schmidt said in an interview with WIRED earlier this year. “They are better than other technical people.”

One of Google’s paradigmatic divas was Andy Rubin, the cofounder of Android, a startup that Google acquired in 2005 for an estimated $50 million. Rubin was known for being territorial, demanding, and impressively inventive. Journalists rarely failed to mention his penchant for oddball devices, including a doorbell at his home with a robotic arm that rang a gong on the interior, and a retinal scanner on the exterior. (“The system makes it easier to deal with former girlfriends,” reported a 2007 New York Times profile, paraphrasing Rubin. “No messy scenes retrieving keys—it’s just a simple database update.”)

As executives saw it, Rubin delivered handsomely for Google. More than anyone else, he ensured the company’s safe and profitable transition into the mobile era. As users moved from desktops to smartphones, Android’s open source operating system acted as a Trojan horse for Google’s consumer apps, most importantly search. By 2013, Android had a billion users, and its success afforded Rubin unbelievable perks, including a $14 million loan with a 1 percent interest rate to buy a beach house in Japan. Page was particularly grateful. At one point he awarded Rubin a $150 million stock grant, before Google’s compensation committee even approved the offer.

Google didn’t invent all the concepts that charged its culture. (“Obligation to dissent,” for example, is borrowed from the management consulting firm McKinsey.) But it tied them together into a coherent, aspirational narrative about engineers as a free-thinking people uniquely capable of reconfiguring the world from first principles. This culture helped recruiting, it helped retention, and it kept the public and regulators rapt with admiration. So what if Google was becoming inordinately powerful? The monopolies of the past engaged in price-gouging and became less inventive. Google’s products stayed free and continued to blow your mind.

Occasionally, of course, a glitch would mar Google’s image as a benign, gravity-defying force in world markets, but even those glitches would sometimes resolve in ways that reinforced the myth. In 2006, eager to tap into a growing market, Google opened an office in Beijing and launched a censored version of its search engine in China. Employees saw this as a clear violation of Google’s principles; they were supposed to make the world’s information universally accessible, not suppress it. The project quickly went awry. Executives had hoped that Google’s presence in China would make the country more open, not closed. Instead, it emboldened the government to demand more concessions. Then, in December 2009, Google discovered a sophisticated cyberattack, originating inside China, designed to access the Gmail accounts of dissidents and human rights activists. Brin, who had been extremely reluctant to enter China all along, convinced Page that they should stop complying with China’s censorship rules and tell the public about the attack.

At the time, Google was run as a triumvirate, with CEO Eric Schmidt playing the role of resident grown-up. Schmidt argued that if Google stopped censoring search results, it would never get back into China. But Page and Brin were unmoved. In January 2010, employees at Google’s office in Beijing learned from a public blog post that the company was pulling back from China. In an emotional teleconference, they told executives they felt abandoned.

The reaction in Mountain View, however, was euphoric. At the TGIF meeting that week, employees burst into thunderous applause and gave Brin a standing ovation. “The legacy of the China decision was a giant dose of goodwill from Googlers around the world,” Schmidt wrote in How Google Works; it reaffirmed the company’s principles “governing how all tough decisions should be made.”

Within five years, however, the costs of that decision—and the limits of Google’s entire formula for success—were starting to become uncomfortably clear. Google was still minting money, but ad revenue growth was slowing, and the cost of hiring engineers and funding R&D was climbing fast. Investors wanted to know what was next, and Google didn’t have a convincing answer. The company was making incredible strides in artificial intelligence, but its growth online was increasingly boxed in by social networking sites. For users hanging out inside Facebook or talking to Alexa, Google’s apps were no longer a click away. And while the company’s founders tinkered with self-driving cars and helium balloons that beamed the internet down to earth, Amazon had built up a huge advantage in an area that should have been Google’s to lose: cloud computing.

Page and Brin had been focused on the cloud for more than a decade. They had built a sublime constellation of data centers around the world to power Google’s own products like Gmail, a service that, for vast swaths of users, had normalized the idea of surrendering your data to a company’s remote servers. But Google’s MO was building products that a billion people would use by default—not closing deals, managing contracts, and customizing infrastructure for other companies.

Amazon had no such compunctions. In 2006 it began marketing a simple but highly effective cloud computing platform to other firms, ultimately offering clients like NASA and Netflix on-demand access to computing power, including storing and processing data. By the time Google finally offered a comparable service called Google Cloud Platform, in 2012, Amazon was already leagues ahead. In April 2015, Bezos revealed that Amazon Web Services had brought in $4.6 billion in revenue the previous year and was on track to out-earn his retail business soon. Google’s earnings call was the same day. The company reported that 90 percent of its revenue was still from online advertising.

Around the same time, Google’s decision to take the moral high road out of China was seeming more and more like a self-isolating move. Across Silicon Valley, tech giants in search of growth were going after China’s then-680 million internet users. Apple had been running an app store in China since 2010. Microsoft’s search engine, Bing, had served censored search results since 2009. Even LinkedIn was there. Meanwhile, Google watched as Chinese handset manufacturers like Xiaomi sold phones that ran on an unofficial version of Android—which meant no Google Search on the homescreen, no Google app store, and no good way to make money off of millions of devices.

The problem wasn’t just that Google was losing a slice of revenue here or there. For Google, these weaknesses in cloud computing and China triggered an existential dread: They meant the company was losing visibility into the way the internet was evolving and what the future would look like.

In 2015, Google embarked on a massive reorganization. Under a new parent company, Alphabet, moon shots and side projects would fall outside of Google, which would be more focused on making money. With Pichai as Google’s new CEO, finding a new footing in China and cloud computing were among the company’s priorities. One of Pichai’s first moves, for instance, was to hire Diane Greene, a cofounder of VMware, a company that helped popularize an early version of cloud computing before the dotcom bust. Greene would head Google’s cloud division. But catching up with Bezos was going to require Google to reorient. Buyers wanted dependability, not wizardry—and they wanted someone in Mountain View to pick up the phone. And to make up for lost time, Greene would need to go after some big clients.

IV.

On August 11, 2017—the day after pandemonium at the Googleplex had prompted Pichai to cancel the town hall to discuss Damore—Google’s executives entertained an unlikely visitor: secretary of defense James Mattis. At headquarters, they met the retired general across a stately conference table, with Brin, Pichai, Greene, and Walker on one side and Mattis’ crew on the other, all situated in the latest-model Aeron chairs in a new shade called “Mineral.”

Mattis was there to talk business. The Pentagon was in the process of rekindling its relationship with Silicon Valley, which had grown up out of military contract work in the 1950s and ’60s. The rise of artificial intelligence made a potential relationship once again seem mutually beneficial. In recent years, partly at the urging of tech executives like Schmidt, the Pentagon had begun pursuing contracts to modernize its digital infrastructure. But before the Pentagon could fully partake in AI and machine learning, the department had to label its stockpiles of data and move it to the cloud.

At the time of Mattis’ visit, Google was reportedly in the process of bidding for a project that would jump-start this transition. It was called the Algorithmic Warfare Cross-Functional Team, otherwise known as Project Maven. The project would involve labeling past drone footage to train a computer vision algorithm so that, once everything was in the cloud, new drone footage could be analyzed automatically. Though it was a relatively small contract, Maven represented an important potential prize for Google. Greene had recently boasted that Google Cloud, which according to analysts had 5 percent market share at the time, could be bigger than Amazon by 2022. Federal contracts offered a quick way to get there. Maven could put Google on the fast track to receive the security clearances it needed to win more lucrative defense and intelligence agency contracts, like Joint Enterprise Defense Infrastructure (JEDI), a massive Pentagon cloud contract that was worth some $10 billion—if any cunning competitor could wrestle it away from Bezos.

Google, Amazon, and IBM were all in the running for the Maven contract. But Google was an especially furtive competitor. The company quietly put in a bid via a contractor, and it prohibited the Pentagon from mentioning Google without prior approval. As Google seemed to close in on winning the contract, executives from the cloud team pondered how a deal with the Pentagon—especially one that could be linked to autonomous weapons—might reflect on Google’s non-evil brand. In September, a few weeks after the meeting with Mattis, they discussed spinning up some positive PR that would focus on the “vanilla cloud technology” aspects of the Maven contract. “Avoid at ALL COSTS any mention or implication of AI,” wrote Fei-Fei Li, a Stanford professor and Google Cloud’s chief scientist for AI. Li had not been involved in the Maven contract, but she was concerned about the hype and confusion building around AI in the public imagination. “Weaponized AI is probably one of the most sensitized topics of AI—if not THE most. This is red meat to the media to find all ways to damage Google.”

Li was right to be concerned about how Maven might be received, but the media was not the only group she had to worry about. When Google won the Maven contract in late September, the company opted not to say anything at all—even to its own employees. But it wasn’t long before Liz Fong-Jones learned about Maven from a group of concerned engineers who had been tapped to lay some of the groundwork for it. They asked her to keep quiet until January while they tried to convince management to change course. Fong-Jones agreed and focused her attention on another fire that was quietly burning inside Google.

V.

Just before 8 am on Monday, January 8, 2018, Praetorian PR, a San Francisco firm founded by a Republican political consultant, sent an email inviting reporters to a “major press conference” with James Damore and his lawyer, Harmeet Dhillon, who serves as the California Republican Party’s national committee member and who “often takes on controversial, high-profile cases.”

“You won’t want to miss it,” the email promised.

Damore sat next to Dhillon as she told a smattering of tech reporters and local news affiliates that her client had filed a class-action lawsuit against Google, alleging discrimination against white people, Asians, men, and conservatives, or any combination thereof. “We don’t normally file 100-page, 200-page complaints,” Dhillon explained, thumbing through a heavy printout of the complaint laid out in front of her. “I didn’t think people were going to believe the outlandish nonsense, so we actually attached screenshots throughout.”

When Damore had filed a set of documents stating an intent to sue Google in December 2017, the documents named Kevin Cernekee as a fellow plaintiff, but that complaint never materialized. In the lawsuit unveiled by Dhillon that day, Cernekee’s name made no appearance. The complaint quoted from Google’s warning letter to him and included details about his interactions with executives and HR as evidence that Google had discriminated against conservatives, but Cernekee’s identity remained mostly invisible to the outside world. At least 169 other Google employees were not so lucky: The screenshots included in the lawsuit revealed dozens of email addresses, profile pictures, and snippets of discussion. Most had been culled from Google’s internal social networks, including an anonymous message board about mental health and a mailing list for gender-nonbinary employees.

Another round of harassment quickly began. Threats poured in, calling for employees critical of Damore to be shot in the head, poisoned, blown up, stalked, doxed, sodomized with a cattle prod, and thrown off a building. As before, many of the employees who got singled out were queer and trans people of color. On forums like 4chan, employee names exposed in the lawsuit were linked to their social media accounts. The personal information of at least three employees was dug up and posted online. Another employee was the subject of a derogatory thread on an online forum called Kiwi Farms, which New York magazine has called the web’s biggest stalker community.

That month, Fong-Jones made another calculated exception to her usual policy of keeping dissent within the Google family. She and 14 other current employees spoke to WIRED about the “dirty war” that was being waged inside Google over the issue of diversity. Most news coverage of Damore’s case amplified its claims that Google was cracking down on conservatives. But the employees argued that something else was going on. HR had become “weaponized,” they said; Googlers on both sides of the battle lines had become adept at working the refs—baiting colleagues into saying things that might violate the company’s code of conduct, then going to human resources to report them. But Googlers on the right were going further, broadcasting snippets of the company’s uncensored brawls to the world, and setting up their colleagues for harassment.

Google’s HR department, for its part, was feeling inundated with policy violations across the spectrum. And according to Fong-Jones and her colleagues, the department was too focused on trying to appear even-handed. Employees had been reprimanded and even fired for criticizing Damore’s memo using terms like “white privilege” and “white boy.” “Promoting harmful stereotypes based on race or gender is prohibited,” Google said in a statement about one such termination.

Soon after speaking to WIRED, Fong-Jones helped prepare a statement explaining to colleagues about why she and the others had gone to the press. She asked coworkers to sign a petition for a safer workplace, demanding better moderation of mailing lists and rules against doxing coworkers. It received 2,600 signatures. But it wasn’t the only employee petition that would make waves that winter.

VI.

By February, word about Maven began to swirl outside the teams of engineers who had first alerted Fong-Jones to the project. So she decided to post about Maven on her internal Google+ page, sharing grave concerns that Google might be helping the US government carry out drone strikes, according to a copy of the post provided to WIRED by another Google engineer. Shortly afterward, the group of engineers posted an internal statement of their own, explaining that they had been told to build an “air gap”—a security measure favored by the Pentagon that physically separates networks to protect sensitive data—and informing coworkers about their efforts to thwart the project.

Fong-Jones hoped to pressure Google’s leaders using entirely internal channels, and at first things seemed to be working. Outraged employees began referring to the engineers as the Group of Nine, and executives felt compelled to respond. On her own Google+ page, Greene tried to assure employees that the contract was only $9 million and merely a “proof of concept.” Within 48 hours, however, Fong-Jones says she received a call from a journalist—Lee Fang of the Intercept—asking her for comment on her Google+ post. Someone had leaked. If word got out, she feared, management would feel backed into a corner. So Fong-Jones contacted Google’s security team herself. Let’s catch the leaker, she said.

Another staffer who was keeping close track of the internal rumblings about Maven was Meredith Whittaker, a program manager inside Google Cloud. In addition to her work at Google, Whittaker also helped run a New York University-affiliated research center called the AI Now Institute, which studies the ethics and social implications of artificial intelligence. On February 28, Whittaker drafted a petition for employees demanding that Pichai cancel the contract. “Google should not be in the business of war,” she wrote.

At TGIF that week, executives were ill-prepared for the blowback. One employee stood up and said she had left her last job because of ethical concerns around defense work. Brin told her that Google was different, because at least here she could ask questions. Whittaker’s petition netted about 500 signatures that night and 1,000 the next day. It became a touchstone for a divisive, months-long internal debate, inflamed by Google’s open culture.

There was no consensus on Maven inside Google’s fractious workforce, which includes former Defense Department researchers, military veterans, and immigrants from countries under US drone surveillance. Even the employee group for veterans was split on the project. But Maven’s opponents were organized in a way that Google hadn’t really seen before. Employees fanned out into different groups. Some scoured Google’s open databases, where they discovered emails that appeared to contradict Greene’s statement about the size of the Pentagon contract; they also found snippets of Python code for computer-vision technology that seemed designed to track human beings and vehicles. Some churned out anti-Maven memes; others kept track of employees who were quitting over the contract. One activist group focused on fact-checking, listing every time they found evidence to contradict the company line. The list got longer and longer.

Greene responded by playing whack-a-mole—locking down mailing lists, deleting documents, or asking employees to redact Google+ posts. Longtime execs were taken aback, and even hurt, by the loss of their employees’ trust, which they had come to take for granted.

In March, Gizmodo broke the Maven story to the public. Inside Google, executives urged employees to reserve judgment; Google’s leaders, they said, were developing a set of AI principles that would govern its business practices and contracts like Maven. Employees should wait for those principles, which would provide the whole company with a basis for discussion.

At another time, such a gesture toward self-regulation might have sufficed. And plenty of employees were eager to change the subject. But the anti-Maven organizers had momentum on their side—and support from an outside labor organization called Tech Workers Coalition. They were also galvanized by a powerful realization: They could reliably summon the rapt attention of the media and the public, which were hungrier than ever for the vicarious thrill of watching someone—finally—hold tech companies accountable.

On April 4, The New York Times published a story about Whittaker’s petition, which by then had gathered more than 3,100 signatures. Four days later, at 10 pm on a Monday night, Whittaker received an email from Greene inviting her to participate in a four-person internal debate about Maven, to be broadcast to Google’s employees from Mountain View that Wednesday.

Whittaker prepared feverishly. She called up colleagues and a friend from the Defense Department. She memorized drone kill stats and read up on defense contracting. On the day she flew out from New York for the event, Whittaker learned that the debate would take place three times during the course of the day, so that employees in other time zones could tune in.

Whittaker should have been outmatched. Greene served as the moderator, and the two pro-Maven panelists were both longtime Google veterans. But they used many of the same arguments that executives had recited in recent TGIFs. The contract is exploratory. Maven merely uses Google’s open source machine-learning software. Better to have Google working on AI than a defense contractor. Maven helps “our” military. Whittaker, building on an analysis she had been rehearsing in Google’s social networks for weeks, argued that the ethical codes surrounding AI were still largely unformed and shouldn’t be defined on the fly in the context of a business relationship with the Pentagon.

After the first panel, Whittaker paced back and forth in the Googleplex parking lot, calling up colleagues to ask how they would have countered her remarks. By the time the last panel ended, Whittaker’s energy was zapped. She ran to her car in the pouring rain, picked up some beer and peanuts at a bodega, and went back to her hotel room. Inside Google, people often look to Memegen to gauge the mood of an unruly workforce. On the day of the town halls, most of the memes were pro-Whittaker.

On May 30, The New York Times published a story about Maven that included the emails Fei-Fei Li had sent to other executives about weaponized AI. At the weekly cloud team meeting two days later, according to details leaked to Gizmodo, Greene announced that Google planned not to renew the Maven contract. The backlash, she reportedly said, had been terrible for the company.

VII.

In early June 2018, Pichai finally published the AI principles that Google had promised its employees. They included a list of four applications of AI that Google would not pursue, including weapons, technologies that gather and use information “for surveillance violating internationally accepted norms,” and technology “whose purpose contravenes widely accepted principles of international law and human rights.”

Two months later, it seemed to many employees that Pichai had already broken those principles. On August 1, a blockbuster story in the Intercept reported that Google was planning to launch a new censored search engine in China. Codenamed Project Dragonfly, the engine would blacklist search terms like “human rights” and “student protest,” and would produce government-controlled results for “air quality.” The service would take the form of an app that Google was prepared to launch in six to nine months. Chinese officials had already seen a demo, according to the Intercept, though Google still needed government approval. The app might even help link a person’s search records to their cell phone, with the information stored on servers in China. (A Google spokesperson tells WIRED that, at the time, it was “impossible to confirm or comment on” what the service “might have looked like—it was too early to say.”)

Once again, an employee backlash set in. The Chinese government censorship that had so disturbed Google staffers in 2010 had only grown bolder and more sophisticated under President Xi Jinping, who was detaining hundreds of thousands of Uighurs and members of other Muslim minority groups in internment camps and deploying the latest in surveillance technology on citizens.

The succession of scandals kept deepening the divide between execs and employee activists. More and more, the latter were questioning the social contract they had lived by. “I went from ‘Oh my god, who leaked that?’ to ‘Oh my god, management did what?!’ ” Fong-Jones says. She started to doubt her past successes with executives. “Perhaps the reason they were willing to listen in the first place was to give up the things that mattered less to them,” she says.

A couple of weeks after the first Intercept story, Pichai answered some questions about Dragonfly, which he described as an “exploratory” program. In his telling, Dragonfly was a project that aligned with Google’s principles, not one that contravened them. “Our stated mission is to organize the world’s information,” he said. “China is one-fifth of the world’s population.” Besides, he said, “I genuinely do believe we have a positive impact when we engage around the world, and I don’t see any reason why that would be different in China.”

Brin also spoke at the meeting, claiming he only learned about Dragonfly because of the “kerfuffle.” (Former executives say this strikes them as implausible.) He also said that “Googlers should feel broadly proud of their work, not feel that it compromises their principles.” Then the meeting stopped abruptly. Someone was live-leaking the event to a New York Times reporter. An executive onstage asked the technicians running the meeting to show everyone what was happening, and a wall of screens behind Brin flashed a tweet from the reporter who was posting Brin’s comments in real time.

With the Dragonfly scandal, employees were in some ways the least of Google’s worries. Washington lawmakers in both parties responded in ominous terms. “It would be very dangerous for Google if they were misrepresenting to American policymakers the extent of their involvement in China or the ramifications of some of their joint ventures,” US senator Mark Warner, a Democrat from Virginia, told WIRED, noting the growing bipartisan concern around China.

Dragonfly left Google particularly vulnerable to criticism from leaders on the right, who painted the company as un-American. Josh Hawley, now a Republican senator from Missouri, said Google was motivated more by money than allegiance to any country. “If that means violating the privacy of American consumers, they’re happy to do it. If that means going to China and partnering on technologies that aid a repressive regime, they’re apparently happy to do it.” In that context, Google’s old slogan no longer applies, Hawley said. “It’s time that we remove the halo.”

VIII.

On October 25, 2018, a story in The New York Times dredged up the dark side of Google’s tolerance for aberrant geniuses. In 2013, the paper reported, a woman who worked for Google had accused Android cofounder Andy Rubin of coercing her to perform oral sex in a hotel room. Google, the story reported, had investigated and found the claim credible but sent Rubin off with a $90 million exit package and a fond farewell.

The story didn’t stop at Rubin. Another high-performing executive, Amit Singhal, the former head of Google Search, was given a multimillion-dollar exit package after a female employee accused him of groping her at an off-site work event. A third, Richard DeVaul, allegedly told a female job candidate that he was in a polyamorous relationship during her interview and invited her to meet him at Burning Man, where he asked if he could give her a massage. Google denied her the job. DeVaul was still working at Google as director of X, the company’s experimental division for ambitious projects.

The Times investigation also noted Google’s overall “permissive culture,” in which top executives—including Brin, Schmidt, and chief legal officer David Drummond—had relationships with female employees. Some of the women claimed they were later pushed out of the company.

Google employees lit up the company’s internal social networks, once again contemplating galling facts about the status of women in Silicon Valley. But this time the discussion was less easily derailed, perhaps because some of the most important exchanges took place on an anonymous mailing list called Expectant New Moms. The group’s 4,000 members knew the stories about Rubin and Singhal—thanks in part to email threads on the list after each executive departed. But Rubin’s $90 million payout felt like a sucker punch. The fact that leaders’ misconduct had been an open secret made it worse. Why had they given so many years of their lives to make these men insanely rich?

At 2:05 pm, Claire Stapleton, then a product marketing manager at YouTube, fired off a message to the group: “absolutely disgusting—all of it, all of them. topple the patriarchy.”

That day Alphabet was already poised to share some mixed financial news. The company was set to report $9.19 billion in profits for the third quarter, thanks in part to Trump’s tax cuts to benefit big business, but missed revenue targets. Now executives scrambled to do damage control with employees before the earnings call. A couple of hours after the Times story was published, Pichai sent a memo assuring employees that Google had reformed its ways. In the past two years, he wrote, 48 people had been terminated for sexual harassment, including 13 at the senior manager level or higher, none of whom received an exit package. Employees were skeptical. If Google was so committed to a safe environment, why was DeVaul still there? (A Google spokesperson said HR investigated the allegation “thoroughly and took appropriate corrective action.”)

That same evening, Page—who was CEO when the claims about Rubin, Singhal, and DeVaul came to light—apologized to employees at a TGIF. “I’ve had to make a lot of decisions that affect people every day, some of them not easy. And, you know, I think certainly there’s ones with the benefit of hindsight I would have made differently,” he said in a prepared statement. Page’s explanation was evasive, but his tone was serious. Brin, on the other hand, made an awkward joke about confidentiality, which sounded to some as if he were blaming the leakers, and seemed irked by the fact that the same question came up over and over again when the executives had nothing new to say. The meeting soon returned to business as usual with a product demo for new Google Photo features.

For Stapleton, the meeting was a huge letdown. Her first job at Google, in 2007, had been to help lead TGIF, including writing talking points for Page and Brin. Now they couldn’t even answer the main question. At 7:58 pm, Stapleton fired off another message to the list suggesting the moms channel their anger into collective action, like maybe a walkout, a strike, or an open letter. “Googler women (and allies) are REALLY ragefueled right now, and I wonder how we can harness that to force some, like, real change.”

As the women dissected the TGIF performance, they also swapped stories about reporting harassment to Google HR, only to watch their abusers receive promotions. The messages were still rolling in at 1 am.

The next morning, Stapleton started a Google Group. “Welcome to ground zero of the google women’s walkout / a day without women (naming/branding tbd).” Just as with the travel ban walkout 21 months earlier, word spread quickly. A group of eight organizers emerged, including Meredith Whittaker, and they got to work planning logistics, hammering out demands, and fine-tuning their message. Stapleton set up a Google form to ask employees why they were walking out. The 350 responses that promptly poured in included personal stories about harassment, discrimination, retaliation, and pay inequity.

From those responses and other internal posts, the organizers smelted together five core demands. They wanted an end to forced arbitration, a process that compelled employees to bring their claims to a private arbiter, paid for by Google, rather than before a judge. They also demanded pay equity and policies that would guarantee more transparency around harassment claims. They scrubbed personal details from the 350 stories and divided them into buckets that mapped to each demand, so the rally organizers would have something to read.

As the plans came together, Google invited Stapleton to join a meeting with three top female executives: Ruth Porat, the company’s chief financial officer; Wojcicki, the CEO of YouTube; and Eileen Naughton, the head of people operations. The invitation was brokered by the head of Women at Google, an employee resource group, who told Stapleton it was an amazing opportunity. Other walkout organizers disagreed. Google was just trying to co-opt their momentum, the organizers felt. Stapleton pushed off the request. “It feels weird to say no to these women who are, like, Illuminati,” she said.

On Tuesday evening, two days before the walkout, Pichai sent another memo: “One thing that’s become clear to me is that our apology at TGIF didn’t come through, and it wasn’t enough.” He acknowledged the protest and told employees they would have the support they needed.

The plan was for each office to walk out at 11:10 am on November 1, 2018. By the time the first images came in from Asia, it was clear that the call hadn’t merely mobilized the lefties in Mountain View. In Singapore, where labor law prohibits workers from marching, employees stood in a cavernous office lobby, somber and listening intently to the speakers. In New York City, employees streamed out from Google’s Eighth Avenue office into a nearby park. Whittaker and Stapleton stood on chairs as they rallied the crowd. Their megaphone was no match for the noise from the West Side Highway, but chants of “Time’s up!” rose above the din.

When it came time for Mountain View to rally, 4,000 Google employees filled the courtyard outside the main café, once again chanting and holding signs, including ones designed by volunteers from Google Creative Lab that said “HAPPY TO QUIT FOR $90M—NO SEXUAL HARASSMENT REQUIRED” in a fresh, sans-serif font. Standing a few yards from where Pichai and Brin had stood during the travel ban walkout in 2017, employees once again shared their stories. A female YouTube employee described being drugged by a coworker at a company event, then being told by HR that she had to stay on the same team. This time, no executive stepped up to the mic. No one chanted their names.

Local TV crews had to report from the edges of campus, but the news choppers overhead had a clear view of “Not OK, Google” and “Time’s Up” written in chalk on the pavement. By then, Stapleton, Whittaker, and the other organizers back in New York had commandeered a booth at a Mexican place in the Meatpacking District. They ordered margaritas and chips and guacamole, updating social media accounts and ducking outside for press interviews. The women were triumphant; they were half asleep. A few days earlier, they’d thought a turnout in the hundreds would be a big deal. When they tallied the total, 20,000 employees had walked out.

In a way, though, Google’s attempts to neutralize the walkout had worked. A sense of catharsis and camaraderie seemed to overshadow any hostility toward management. Even some supporters felt that the walkout was more about the act than the asks. On November 8, 2018, a week after the protest, Pichai sent a memo to his employees—simultaneously published on Google’s blog—announcing that Google would no longer require arbitration on sexual harassment claims. But it would still prohibit class actions. The change brought Google in line with policy tweaks that Microsoft, and even Uber, had made within the past year.

IX.

Google’s response followed a familiar cycle: internal opposition and bad press, followed by incremental change. But the Women’s Walkout appears to have perturbed executives in a way that the protests against Dragonfly and Maven had not. In the months that followed, pushback seemed to spill over into the organizers’ day jobs. In December, Whittaker was told she would have to leave the Google Cloud organization, where she had worked for three years. A few weeks after that, Stapleton claims she was told that her role at YouTube would be “restructured” and she would lose half her reports and responsibilities. (A Google spokesperson says no changes were made to Stapleton’s role.)

Fong-Jones, for her part, was burning out. It had been a long year and a half since Damore’s memo leaked. In early January 2019, she submitted her resignation, but she turned even that into a last-ditch effort to reform Google; she told executives she would reconsider if the company put an employee on its corporate board. Instead, Google HR tried to get her to leave before her notice period was up, and she filed a retaliation claim. Google investigated and determined it was unfounded.

That month, Google also tightened the reins on TGIF. Brin and Page stopped showing up. Employees could access video recordings for only a week after the meeting, rather than for years. The company nixed live questions, which Google claimed was more fair to employees in different time zones. (“We’re a global company and want to make sure we’re answering questions from employees around the world,” a spokesperson says.) TGIF’s transformation from candid conversation to press conference was pretty much complete.

The company’s internal social networks were quieter too. Cernekee had been fired in June 2018 for violating multiple company policies, including using a personal device to download company information. As part of his ongoing legal case, according to people familiar with the matter, he had to give back 20,000 pages of internal documents, some of them confidential. Cernekee disputes this.

(For weeks, as WIRED was preparing this story for publication, Cernekee’s lawyer tried to dissuade the magazine from disclosing his identity. Six days after this article went to press—but before it appeared on newsstands—an interview with Cernekee appeared in The Wall Street Journal. He presented himself as a whistle-blower and mainstream Republican who had been bullied and eventually fired for his beliefs. In subsequent interviews with Tucker Carlson and on Fox & Friends, Cernekee said he believed that Google would try to sway the 2020 election—a claim that inspired a tweet name-dropping the engineer from President Trump himself.)

Damore’s class-action lawsuit, meanwhile, was still proceeding, albeit without Damore. In October 2018 his claims were moved to arbitration, while claims from two conservatives who allege they were denied positions for political reasons are proceeding in court.

A couple of months later, Pichai was called up to answer questions about Dragonfly at a House Judiciary Committee hearing. “Right now,” he said, “there are no plans for us to launch a search product in China.” That wouldn’t be his last trip to Washington. After two years of employee revolts, culture-warring, and accusations, the furor that had surrounded Facebook for the previous three years now finally seemed to be aimed at Google with full force. Within a three-week span in March, Senator Hawley banged the drum on amending laws that granted platforms immunity from liability for moderating their platforms; Democratic senator and presidential candidate Elizabeth Warren published a plan to break up Big Tech; and the chair of the Joint Chiefs of Staff testified that Google’s effort to court China “indirectly benefits the Chinese military,” even as it snubbed working with the Pentagon. A couple of days later, President Trump tweeted, “Google is helping China and their military, but not the U.S. Terrible!”

That spring, many of Google’s efforts to stave off scrutiny seemed to hasten its arrival. In March, Google announced that it had formed an AI ethics council composed of external advisers. It included Kay Coles James, president of the Heritage Foundation, who lacked any discernible expertise in artificial intelligence and who had recently expressed anti-trans and anti-immigrant views. Some employees were appalled; an internal petition to remove James quickly gained 2,500 signatures. Breitbart and the Daily Caller posted the names of petition organizers, including Whittaker, and leaked internal messages from a mailing list. One ethics council member quit in the midst of the uproar. When Google learned that another member was planning on defecting, the company disbanded the council—nine days after it launched. To the outside world, it looked as if Google had capitulated to employee protests. Conservative critics descended. In The Washington Post, James said Google was not upholding its bargain with the right. “How can Google now expect conservatives to defend it against anti-business policies from the left that might threaten its very existence?” she asked.

While Whittaker was helping to lead the charge against the ethics council, she continued to tussle with management over her job. Stapleton, who had been told multiple times that she was a “rising leader” in YouTube marketing, says she was also struggling to hold on to the responsibilities she’d held before the walkout. When the two women heard that a third organizer had also been denied a transfer, they posted an open letter on the walkout mailing list on a Monday in late April, reporting to their coworkers that Google was retaliating against them. They invited their colleagues to fight back against retaliation at an employee town hall meeting that Friday, to be livestreamed at Google offices around the globe.

That week, managers emailed the entire marketing and cloud departments, denying the women’s claims. On the morning of the town hall, Lorraine Twohill, the head of marketing at Google, also sent a department-wide email saying Stapleton’s claims were false. “Over the last several weeks, I have spent a lot of time talking to everyone involved, trying to understand and empathize with the situation,” she wrote. Stapleton says Twohill never asked her about the incidents surrounding her claims of retaliation.

After that, Stapleton saw no future for herself at YouTube. She handed in her resignation three weeks later. In mid-July, Whittaker resigned too. The next day Google happened to announce definitively that Project Dragonfly was dead. By then, none of the major organizers of the protests that had shaken Google over the past two years were left at the company.

But that didn’t mean things would go back to normal at Google. Over the past three years, the structures that once allowed executives and internal activists to hash out tensions had badly eroded. In their place was a new machinery that the company’s activists on the left had built up, one that skillfully leveraged media attention and drew on traditional organizing tactics. Dissent was no longer a family affair. And on the right, meanwhile, the pipeline of leaks running through Google’s walls was still going as strong as ever.

Late this June, Project Veritas, a right-wing outlet specializing in stings and exposés, published a slew of leaked documents and snippets of hidden-camera footage from inside Google. One of the items it posted, as if evidence of Google’s supposed bias, was a “Beginner’s Guide to Protesting” that Google employees had drawn up around the time of the travel ban walkout back in 2017. Next to the document was a message and a link. “Do you work in Big Tech?” it said. “Project Veritas would love to hear from you.”

NITASHA TIKU (@nitashatiku) is a senior writer at WIRED.

This article appears in the September issue. Subscribe now.

Let us know what you think about this article. Submit a letter to the editor at mail@wired.com.

When you buy something using the retail links in our stories, we may earn a small affiliate commission. Read more about how this works.

More Great WIRED Stories

• The weird, dark history of 8chan and its founder
• 8 ways overseas drug manufacturers dupe the FDA
• A Boeing code leak exposes security flaws deep in a 787
• The terrible anxiety of location sharing apps
• Seeing through Silicon Valley’s shameless “disruption”
• ??‍♀️ Want the best tools to get healthy? Check out our Gear team’s picks for the best fitness trackers, running gear (including shoes and socks), and best headphones.
• ? Get even more of our inside scoops with our weekly Backchannel newsletter