What Careless People Teaches Us About the Tech Lobby’s Playbook

Sarah Wynn-Williams’ new book shows us that “It didn’t have to be this way”

Intro from Jon Haidt:

We describe a book as “riveting” when we feel our attention firmly attached to it, as though we can’t pull ourselves away. That’s a powerful term of praise, especially in the 2020s when teens as well as adults are finding it harder to read books or to think, reason, or generally pay attention to one thing for longer than five minutes. I’m now reading Careless People, the new book by Meta whistleblower Sarah Wynn-Williams, and it is a riveting account of life and policy inside one of the corporations that (I believe) has done the most (along with TikTok) to fracture and frack humanity’s collective attention.

Wynn-Williams joined Facebook in 2011, full of youthful idealism about Facebook’s mission. She convinced the leadership that they needed someone like her to handle government and political relationships. She then created that new position and became Facebook’s first international public policy lead. She spent the next six years helping Facebook to develop useful relationships with governments around the world. Her idealism quickly faded as her work became an unending series of crises caused, in many cases, by the extraordinary carelessness of Facebook’s leadership. “Move fast and break things” really was their credo. Growth at any cost, regardless of who gets hurt along the way.

To give the example most relevant to readers of After Babel: Wynn-Williams says that Facebook was appealing to advertisers by telling them that Facebook can microtarget ads based not only on demographics but based on a users’ emotional state as well. She writes:

In April 2017, a confidential document is leaked that reveals Facebook is offering advertisers the opportunity to target thirteen-to-seventeen-year-olds across its platforms, including Instagram, during moments of psychological vulnerability when they feel “worthless,” “insecure,” “stressed,” “defeated,” “anxious,” “stupid,” “useless,” and “like a failure.” Or to target them when they’re worried about their bodies and thinking of losing weight. Basically, when a teen is in a fragile emotional state.

A reporter in Australia had obtained an internal Facebook document and he reached out to Facebook for a comment before publishing his story. That’s when Wynn-Williams learns about the emotional targeting because she’s put on a team to manage the PR fallout. She and the team consider putting out a statement along the lines of “We take this very seriously and are taking every effort to remedy the situation,” but the team discovers that this emotional targeting is widespread, and that Facebook is doing nothing to discourage it. Wynn-Williams gives one example she learned of: Facebook works with a beauty product company to track whenever a teen girl deletes a selfie, which is presumably a moment when she was not satisfied with how she looks. At that very moment, Facebook serves them an ad for beauty products.

Careless People is not primarily about what Facebook and its various platforms did to children. It is primarily the story of how Facebook used its money, ambition, platform, and vast carelessness to manipulate the governments, regulators, and public opinion around the world to fend off any kind of policy that would have forced them to care about the millions of people they were harming in their all-out race for growth.

Careless People is a mega bestseller in its first few weeks in part because it is a well-written bombshell, but also because Meta tried to block the publication of the book and has obtained a court order barring Wynn-Williams from speaking publicly about the book (claiming that speaking badly of the company would violate the agreement she signed when she left the company). But in a lovely example of the Streisand Effect, the publicity around Meta’s efforts to silence Wynn-Williams seems to be driving more people to read her book.

Image. Careless People

To draw lessons from Careless People, I turn to, Casey Mock, who has his own Substack, Tomorrow’s Mess, and brings years of experience in public policy, including four years at Amazon on the company’s public policy team; with the Center for Humane Technology, where he led policy and public affairs; and prior to that, working for two U.S. governors, one from each party. So he’s worked on many sides of the policymaking equation.

Casey’s essay is not a review or summary of Careless People. Rather, Casey brilliantly extracts the Meta playbook for avoiding regulation from the revelations in Careless People. The playbook has similarities to the one used by the tobacco lobby, as Gaia Bernstein showed in a previous After Babel post.

It is vital that the world understand this playbook for it reveals how Meta (and other tech companies allied with it) are able to block legislation in so many countries. It explains how Meta was able to stop the Kids Online Safety Act (KOSA) last fall, even after it had passed the Senate with an overwhelming bipartisan endorsement (a vote of 91 to 3) and had strong bipartisan support in the House of Representatives.

Casey lays it all out below, drawing on Wynn-Williams. Please read Casey’s analysis, and then go read Wynn-Williams riveting new book.

– Jon

---

What Careless People Teaches us About the Tech Lobby’s Playbook

Source: Shutterstock

During the January 2024 Senate Judiciary hearing on social media's impact on children, Senator Josh Hawley publicly admonished Mark Zuckerberg, telling him he should take the opportunity to apologize to the parents in the room. Zuckerberg turned to the crowd and said:

I'm sorry for everything you have all been through. No one should go through the things that your families have suffered and this is why we invest so much and we are going to continue doing industry-wide efforts to make sure no one has to go through the things your families have had to suffer

This moment appeared to signal a turning point in tech accountability. Yet within months, the Kids Online Safety Act (KOSA) died in the House without a vote, while the TikTok ban—which conveniently eliminated Meta's biggest competitor—sailed through with bipartisan support.

Meta’s governmental strategy and influence is now clearer than ever, thanks to Sarah Wynn Williams’s recently published memoir, Careless People. In her account of the culture of callousness, greed, unaccountability, and nepotism among Meta’s leadership – including Zuckerberg, former COO Sheryl Sandberg, and current President of Global Affairs Joel Kaplan – she details the company’s well-honed political playbook. Without any apparent sense of irony for a company that itself now refuses to fact-check, Meta is engaged in a legal and PR campaign to silence Wynn-Williams, claiming that the book contains fabrications and was not properly fact-checked. The publisher, Macmillan, stands by the book.

As someone with over a decade of experience in tech policy—including four years on Amazon's public policy team—I found Careless People compelling and credible. Her insider’s account makes clear how Meta, despite widespread public awareness (thanks in part to Frances Haugen’s testimony), has repeatedly evaded meaningful accountability, even as the company facilitated genocide in Myanmar, created political chaos in the US and Europe, and manipulated the emotional vulnerabilities of teenagers for profit. Wynn-William’s revelations provide a behind-the-scenes look into a regulatory avoidance strategy I know well, because it’s the same political playbook Amazon uses, as well as Google, Microsoft, and other large tech and social media companies.

Careless People has pulled back the curtain on Meta's sophisticated political playbook—one that's shared across technology and social media companies. In the rest of this post, I’ll share four aspects of this strategy that I recognized. Understanding these tactics matters for anyone concerned about both tech monopolies and children's safety online.

1. Policymakers and governments are mere nuisances to be managed

This is less a tactic than an orientation, but undergirds each of the other tactics described below, and ultimately derives from tech culture’s scorn for institutions.

Wynn-Williams relates several anecdotes about the instincts of Meta’s senior leadership in dealing with policymakers, ranging from chronic disrespect for the time of heads of state by showing up late to meetings and refusing to schedule anything before noon – which made me cringe in empathy with Wynn-Williams – to open hostility to the very idea that Facebook could even be subject to the laws of countries like Brazil, South Korea, or China. Meta’s evolution in policy rhymes with the familiar tales about tech and DC – from Microsoft to Google, start-up tech companies historically have seen themselves as special and above petty politics, before “maturing” into a posture of leveraging aggressive global lobbying as a way to fight their competition and, overwhelmingly, antitrust enforcement.

Wynn-Williams, who worked as a diplomat before joining Meta, was clearly baffled by this attitude, both when she first encountered it and as it escalated. Mark Zuckerberg’s famously snide response to Senator Orrin Hatch’s question about how the company intends to continue offering its product for free – “Senator, we run ads” – was less than a decade ago, but an entire generation of tech companies have matured in the wake of that moment. I encountered it myself at Amazon, best illustrated by when executives disastrously testified to the New York City Council about its plan to open a second headquarters in Long Island City, disdainful that they even be subject to a democratic process to receive taxpayer subsidies.

2. Investments buy policy goodwill

Wynn-Williams describes how Meta leveraged the prospect of buildinga data center—that eventually went to Iowa—as a means to influence policy changes in Canada. She also recounts a ‘wink-wink’ negotiation between Sheryl Sandberg and the Prime Minister of Ireland for Meta, in which Meta secured continued favorable tax treatment in exchange for maintaining high-paying positions in Dublin.

This is a tried-and-true tactic that’s older than tech, but tech has mastered it: from Meta’s timely data center announcement in Louisiana last year, to the Foxconn factory in Wisconsin,¹ to the SoftBank/Oracle/OpenAI Stargate press conference earlier this year,² tech companies know that policymakers love to attend ribbon-cutting ceremonies to inaugurate showy investments and demonstrate to their constituents that they are bringing jobs to their districts. And the fact that these investments buy goodwill with policymakers knows no party or ideology. To name one example, it negated the political impact of Senator Elizabeth Warren’s criticisms of Amazon when she attended an opening ceremony for the new warehouse in Fall River, Massachusetts.

3. Use “third party validators” as pawns

Careless People includes an anecdote how Facebook funded groups like Human Rights Watch, Freedom House, and Reporters Without Borders” in order to neutralize organizations that might criticize Facebook” for the choices it made in entering the China market, and later Wynn-Williams shares a quote from the former managing editor of Snopes: “They’ve essentially used us for crisis PR . . . They are more interested in making themselves look good and passing the buck.”

This is perhaps Meta and tech’s most effective strategy: leveraging “third-party validators,” which is an internal tech lobbying euphemism for “pawn,” or, in some cases, “useful idiot.” Meta has forged partnerships with valuable third-party validators beyond those Wynn-Williams names, who have clearly been selected to help Meta launder its reputation around kids’ safety issues. When groups that would seem unlikely allies of tech companies don’t show up in support of (or in some cases, even testify against) regulation of the internet, social media, or phones in schools, they often tip the scales decisively in tech’s favor, sometimes without tech’s lobbyists even having to show up themselves.³

I can share two examples of this in action. In 2023, the Minnesota Star-Tribune lobbied against state legislation that would make online spaces safer for kids, even though the legislation included a specific carve-out for news organizations. The publisher of the Star Tribune is Steve Grove, who was not only an effective lobbyist with legislators because he had just left a position in Governor Tim Walz’s cabinet, but he also was a friendly face for the tech companies opposing the legislation, in that he is a former Googler who helped found Google News. The very fact that Grove and the Star Tribune got involved was decisive in stopping that legislation, with Grove never having been questioned on the record why the legislation would have been an imposition on the Star Tribune.

And perhaps the most egregious example of this tactic is from Amazon, which paid a firm $30,000 to create from whole cloth a new disability rights nonprofit just to oppose privacy legislation in California.

4. Dissemble, distract, and outlast

“We’ve been working on this problem for a long time and we take this responsibility seriously. We’ve made significant progress but there is more work to be done.”

This is Wynn-Williams quoting a post Zuckerberg made after he first realized that Facebook played a role in Trump’s first election, but you could copy/paste this response in place of Meta’s response to any crisis or incident it has played a role in and it would not be materially different.

This isn’t substantive; it’s spin. Over the last fifteen years, tech companies from Meta to Google to Amazon became so big that they could hire the best legal counsel and corporate PR that money can buy. But a consequence of ever-greater spending on corporate PR is an evolution from what once was admirably cheeky and refreshingly direct communications with policymakers and the public to insincere statements that say nothing. A close cousin of this approach is announcing changes to a product proactively that then later get quietly moth-balled or out-sourced when either a competitor takes market share by engaging in riskier behavior or it becomes a financially unsustainable practice. Changes made by CharacterAI following high-profile lawsuits this past fall fit this fact-pattern.

But while we get nothing of substance from Meta’s spokespeople, they and other companies use mouthpieces that they keep at arms’ length. The most sinister and increasingly well-funded of these mouthpieces is NetChoice – bankrolled by Amazon, Google, Meta, and other prominent tech and social media companies – which opposes any and all social media and smartphone regulation, proposes no alternative workable solutions (unless you count advising that parents worried about their kids on screens and social media should go to a “genius bar for parents” or that parents should just “go upstairs and [take] the f—ing phone away”).

NetChoice is part of an alphabet soup of groups that hold themselves out as standing up for values many people support – free speech, an open internet, etc. – but only take the stances their paying members ask them to take. I know this, because while at Amazon, I told the lobbyists for organizations like NetChoice, TechNet, and others, exactly what legislation to oppose and how to oppose it, and it was often helpful to have these groups serve as my bulldog while I cultivated a more friendly approach that protected the brand.

The combined impact of groups like NetChoice and the leveraging of third-party validators described above is threefold. First, it creates the illusion of a chorus of diverse voices singing in harmony in opposition to a new policy, even though they are very well-coordinated and actually represent the same handful of companies, even though they typically try to hide who they represent. Sadly, lawmakers routinely fall for this – especially state lawmakers, who often don’t have staff and continue to work full time jobs in addition to serving as state legislators, and thus don’t have the time to do deep background research on who is lobbying them and why. Companies know this and exploit it.

Second, it provides an opportunity for the companies to protect their brands while still working to prevent any meaningful new policies from advancing. One example of this is Snap’s public support for KOSA; they have not explained how they reconcile that half-hearted public support with the fact that they have continued to pay membership dues to organizations like NetChoice. Amazon publicly supported a federal online sales tax starting more than a decade ago while it was working behind the scenes to undermine that in state capitals by asking states to legislate instead on terms they dictated, in part by leveraging capital investment.

And third, good faith lawmakers slow the process down (as they should) to try to reconcile what appears to be a bunch of different stakeholders at the negotiating table. Wynn-Williams illustrates this powerfully with a scene in which, after the 2016 presidential election, heads of state approach Zuckerberg as supplicants, and she realizes he will outlast most of them. This problem of lost momentum is particularly acute at the US state level, where state legislators must not only pass a balanced budget, but also have limited time to meet and consider legislation, so tech companies know they can easily run out the clock.

Thankfully, some policymakers are beginning to confront these tactics: Vermont’s legislature held an unprecedented, bipartisan hearing last year, in which Republican and Democrat lawmakers from a diverse set of states compared notes about how the tech lobby had worked to deceive them. But dissemble, deflect, and outlast still works, as last year’s KOSA effort shows. Not six weeks after Zuckerberg's disastrous Senate hearing did the TikTok bill get introduced, with a wave of outside “national security experts” descending on Capitol Hill to scare lawmakers into pursuing just TikTok, rather than legislation that would address not only TikTok's most addictive features but would also impact Meta and other companies. Meta's spokespeople conspicuously denied lobbying on the TikTok bill—a curious claim, given that their lobbying disclosures for the quarter mention lobbying specifically on China and online safety legislation, and that I’ve personally heard from multiple knowledgeable and connected DC contacts that Meta was ultimately behind the TikTok push.

How to Fix “Smashed Up Things”

In Careless People, Wynn-Williams compellingly paints Zuckerberg, Sandberg, Joel Kaplan, and other Meta executives as exactly what the title of the book refers to: the modern equivalents of Tom and Daisy from The Great Gatsby, who “smashed up things and creatures and then retreated back into their money or their vast carelessness,” as the book’s epigraph quotes.

The success of the tactics I’ve outlined above contribute to a larger problem Wynn-Williams identifies, which is the impunity of the tech elite. As she writes, “Facebook is an elite product, born in an elite college, fronted by elite Harvard grads who show up for other elite Harvard grads, who are decision makers in all sorts of places.” Wynn-Williams realizes very late in her tenure that these people will not only never accept her into their circle on her terms, but that there’s a point to the nepotistic hiring of friends at the top echelons of the company: it allows them to circle the wagons and protect their own. This is not unique to Meta; at Amazon, one policy executive hired extensively from his social circle, allowing him to more effectively wield control in the office, much like how Sheryl Sandberg’s and Joel Kaplan’s social circles determined what was possible for Wynn-Williams at Meta. Tech companies like Amazon, Airbnb, Apple, and others conspicuously hired high-profile Obama administration alumni in order to merge powerful social circles, like two aristocratic families marrying each other to entrench their power.

This entrenched power is part of a broader societal problem -- heavily concentrated in Silicon Valley -- of impunity. Section 230 protections and the aggressive lobbying efforts described above have systematically dismantled legal accountability frameworks – and prevented new ones from emerging – that might otherwise check the behavior of tech executives. Simultaneously, the tech industry's disruption of traditional media economics has hollowed out the very institutions that once served as independent watchdogs capable of exposing corruption and calling powerful figures to account. As these companies accumulate unprecedented wealth and influence, the gap between Silicon Valley elites and everyone else continues to widen, giving them disproportionate power to shape policy in Washington and state capitals across the country.

Near the end of the book, Wynn-Williams writes:

It really didn’t have to be this way. I can’t state that strongly enough. If I had to sum up what seven years of watching the people running this massive global enterprise taught me, it’s that something else was possible. They really could have chosen to do it all differently and fix so much of what’s been destructive about Facebook.

She’s right, but at the same time, all is not lost. While tech elites have constructed formidable barriers against accountability, individuals still possess meaningful agency to reclaim control. .

So, what can we do?

First, we can change our personal and family habits by following the four norms outlined in The Anxious Generation: delaying smartphone ownership until high school, keeping social media out of childhood, phone-free schools, and fostering independence through gradually increasing responsibilities and freedoms in the real world. As more families, schools, and communities adopt these practices, tech companies' grip on our economy, media, and politics naturally will weaken.

Second, we must support policies that hold tech companies accountable for what they break. "Move fast and break things" might seem fair to Mark Zuckerberg, but most of us were raised on the Pottery Barn Rule: you break it, you buy it. Companies that ignore safety warnings from their own internal experts and knowingly put dangerous products on the market should face consequences in court. Currently, this is extraordinarily difficult due to existing laws that shield tech companies from liability, even for behavior that would clearly be illegal in the physical world.

Impunity is neither inevitable nor permanent. By reclaiming childhood, understanding the playbook, and demanding legal accountability, we can begin to correct the power imbalance that has allowed a small group of executives to reshape society while avoiding responsibility for the damage they cause. The first step is recognizing that the power of these careless people, while extensive, is not absolute. Thank you, Sarah Wynn-Williams, for exposing the playbook used by Meta’s leadership, and for showing us that it doesn’t have to be this way.

1. Initially announced as a $10 billion flatscreen television manufacturing complex with 13,000 jobs and $3 billion in public subsidies in 2017, and heavily promoted by Governor Scott Walker and President Trump, the Foxconn factory never made a single television and ultimately created fewer than 1500 jobs.

2. Stargate, a joint venture between OpenAI, SoftBank, Oracle, and MGX, was formally announced by President Trump in January 2025, with a planned $500 billion investment by 2029. Elon Musk, among others, has questioned whether the full amount of claimed funding actually exists for this project.

3. I want to be clear that these groups themselves are typically acting in good faith in representing their constituencies, and it’s understandable in cases where such cash-strapped groups accept money or resources from tech companies.