A review of Nick Clegg’s How to Save the Internet: The Threat to Global Connection in the Age of AI and Political Conflict (Bodley Head/Penguin, September 2025)

In 2018, Mark Zuckerberg, the co-founder and chief executive of Facebook, hired Nick Clegg, a former British deputy prime minister, to help manage a growing array of public relations controversies brewing in the United States and Europe. Politicians and the wider public were questioning Zuckerberg’s carefully crafted utopian image of Facebook as a venue for wholesome personal connection. It was becoming evident that, in addition to whatever constructive uses they offer, Facebook, its sister app Instagram, and other mass social media platforms could exacerbate the spread of divisive falsehoods (as occurred when Russian operatives interfered in the 2016 US presidential election), facilitate terrorism, and compromise users’ privacy.

Clegg spent six-plus years at the company now called Meta doing damage control. He left in January 2025, having been promoted to president for global affairs. In one sense, Clegg succeeded spectacularly. Despite a never-ending series of public embarrassments and exposés — as well as a hugely expensive failed effort to diversify into virtual reality — Meta cruised to phenomenal profits and remains the world’s largest social media corporation.

But Clegg’s legacy, which he shares, of course, with his former boss, is one of profound dishonesty, hypocrisy, and human damage — damage not just to many of the billions of individuals who use Meta’s platforms, but also to the societies around the world where the company has operated.

In a new book entitled, "How to Save the Internet: The Threat to Global Connection in the Age of AI and Political Conflict," Clegg recounts his accomplishments at Meta, such as they are, and attempts, unconvincingly, to portray himself as a guerrilla reformer. Apart from a few exceptional passages, the work is a tedious, self-serving survey of practically every issue on the digital agenda. Clegg avoids what could have been the most useful and interesting approach to his topic — namely, providing an insider’s perspective on how decisions were made during a crucial period at a corporate Leviathan. Instead, assuming he signed a nondisclosure agreement, he appears to have embraced it with a zeal that starves the reader’s curiosity.

And yet, I’m going to try to squeeze some value out of having soldiered through the 320 pages that are "How to Save the Internet," if for no other reason than the book neatly illustrates the sophistry and misdirection that Clegg employed during his Meta career — and which I tracked during that span as a tech policy analyst at the Center for Business and Human Rights at New York University’s Stern School of Business. The book’s lessons are pretty much the opposite of what its author claims, but the methods that Meta has used to maintain its power and influence demand attention.

Snapshots of Zuckerberg

It’s difficult to tell what Clegg genuinely thinks about anything. Start with Zuckerberg. Early on, Clegg offers bland encomiums. The mogul is a “visionary innovator” who possesses “endless curiosity” and “indefatigable competitiveness.” He “has the humility, drive, and appetite to keep learning.” Strangely, over the rest of the book, Clegg offers only two memorable snapshots of Zuckerberg and both, perhaps inadvertently, make the CEO seem like a doofus.

In one, Clegg recounts a meeting he brokered in 2024 between his boss and French President Emmanuel Macron, which included a presentation of Meta’s work on artificial intelligence. “The president’s faith in AI can hardly have been boosted,” Clegg writes, “when the AI-powered Ray-Ban Meta glasses Mark showed off mistook a Pierre Soulages painting for a solar panel.” That’s it; end of anecdote. No discussion of the persistent weakness of Meta’s consumer AI products. As it happens, Zuckerberg suffered another cringeworthy Ray-Ban AI glasses failure on September 25, 2025, at a widely covered Meta Connect live demonstration.

The book’s other Zuckerberg scene elaborates on the absurd back-and-forth the CEO had in 2023 with rival tech tycoon Elon Musk over a mixed martial arts cage match that never happened. Clegg reports that Musk chickened out. But the author doesn’t leave it there. He goes on to describe an off-site management retreat featuring a session in Zuckerberg’s purpose-built MMA facility:

We all paired off to practice some moves under the watchful eyes of Mark’s professional instructors, which meant I found myself wrestling with my then-deputy, Joel Kaplan. At one point, this involved a maneuver apparently known as the ‘Domination Mount,’ in which Joel straddled me and we grappled awkwardly in a way that was, let’s just say, too close for comfort. It was corporate bonding taken to a whole new level. Joel later jokingly confessed that he had considered reporting it to our then-head of HR, Lori Goler, but when he looked up to find her, he saw that she had Mark Zuckerberg in a chokehold.

I recount these two episodes to underscore my impression that Clegg actually harbors a certain personal disdain for Zuckerberg. That’s intriguing, and I wish the book gave me more to say about it.

On to weightier concerns.

A straw man fallacy

Mass social media use has contributed to a variety of harms, not least by heightening political polarization, facilitating political violence, coarsening civic discourse, and undermining the mental health and eroding the attention spans of many — if not necessarily all — adolescent users. Clegg spends much of his book seeking to obfuscate these well-documented realities by indulging in the methods that characterized his career at Meta: knocking down straw men and other forms of verbal gymnastics.

Note the careful way I phrased my description of harms associated with social media use: The technology has contributed to…heightened…facilitated…affected many, if not necessarily all. This is the way that serious and credible critiques of social media are worded. Rolling out a classic straw man fallacy, Clegg consistently does battle with a distorted version of the indictment, one that alleges crudely that social media is the original, primary, or even sole cause of the deleterious consequences in question.

For example, after referring to a series of social science studies for which Meta provided outside academics with data, he declares that there is “a growing body of research showing there is little evidence that key features of Meta’s platforms alone cause harmful affective polarization.” The key word in that sentence is alone, which robs the conclusion of any significance. That’s because only a fool or a polemicist would claim that social media alone causes a phenomenon as complicated as affective polarization.

When social scientists use the term “affective polarization,” they mean a form of partisan hostility characterized by seeing one’s opponents as not only wrong on important issues, but also abhorrent and dangerous. A range of factors have contributed to a rise in affective polarization in the US since the 1960s. These include substantive political conflicts such as the backlash against civil rights campaigns on behalf of nonwhite and LGBTQ+ Americans and the realignment of the formerly heterogeneous major political parties into a solidly conservative Republican Party concentrated in the South and a distinctly liberal Democratic Party with strongholds on the East and West Coasts.

The evolution of traditional media industries has also heightened affective polarization, especially the rise of conservative talk radio in the late 1980s and hyper-partisan cable television in the late 1990s. More recently, the uniquely vituperative strategy of President Donald Trump — “I hate my opponent[s],” he said recently — has amplified levels of sectarian strife.

Building on and mingling with all of these trends, mass use of social media has exacerbated partisan animosity. As far back as October 2020, a group of 15 researchers published an article in Science concluding that “social media companies like Facebook and Twitter have played an influential role in political discourse, intensifying political sectarianism.” Reinforcing the point, a separate quintet of researchers summed up their review of the empirical evidence in an August 2021 article in Trends in Cognitive Sciences: “Although social media is unlikely to be the main driver of polarization,” they concluded, “we posit that it is often a key facilitator.”

Clegg disingenuously ignores this nuanced perspective, preferring to play word games. Only a small fraternity of Clegg watchers would know that the author’s sophistry-between-book-covers cannibalizes earlier corporate blog posts on these topics, as well as congressional testimony delivered by Mark Zuckerberg that sought to distance social media use from the January 6, 2021 insurrection. In an August 2023 article for Tech Policy Press, the site’s founder and editor, Justin Hendrix, and I revealed not only Clegg’s reliance on straw man deception but also his dramatic oversimplification of the findings of the research that Meta had arranged and which Clegg cites in his book.

In brief, the studies found that various interventions affecting how subjects used social media didn’t significantly reduce polarization levels. But as Hendrix and I explained, the interventions — such as replacing Facebook’s content-ranking algorithm with reverse-chronological ranking — were so short-lived and mingled with other influences as to raise the question of whether the studies had proven much of anything. They certainly didn’t exonerate Meta in the way Clegg claimed at the time or in his book.

Behind and beyond the polarization problem

Having blurred the connection between social media and polarization, Clegg also obfuscates the engine powering online strife. This mechanism is well understood, according to the 15-scholar collective that published the Science article. “Social media technology employs popularity-based algorithms that tailor content to maximize user engagement,” the co-authors wrote. Maximizing engagement increases affective polarization, they added, especially within “homogeneous networks,” or groupings of like-thinking users. This is “in part because of the contagious power of content that elicits sectarian fear or indignation.”

As I’ve written previously, “Social media companies do not seek to boost user engagement because they want to intensify polarization. They do so because the amount of time users spend on a platform, liking, sharing, and retweeting, is also the amount of time they spend looking at the paid advertising that makes the major platforms so lucrative. Content that elicits partisan fear or indignation is particularly contagious and helps fuel this advertising business model.”

Clegg’s account of this aspect of social media technology is so incoherent as to suggest intentional dishonesty. First, he acknowledges that “when people’s experience using Facebook or Instagram are full of negative emotions, they may be more likely to engage with the content they see, and as a result they may well spend more time on the app in the short term.” Then, offering no evidence whatsoever, he merely asserts: “But it’s not what keeps them coming back day after day, month after month. For that to happen, they have to find the experience useful and meaningful.”

This entirely unsubstantiated distinction between the “short-term” promotion of potentially harmful negative content versus the supposed longer-term appeal of constructive content flies in the face of a raft of studies compiled by researcher Nathan Witkin, showing, in his words, that “our attention on social media gravitates toward content that is more emotionally extreme.” The studies Witkin collected “have found that posts that are sad, fearful, uncivil, morally and emotionally indignant,hostile towards out-groups, and negatively valenced in general spread much further than their more neutral counterparts.”

In a carefully reasoned essay on Substack entitled, “The Case Against Social Media Is Stronger Than You Think,” Witkin argues for a focus broader than affective polarization:

The last decade and a half has seen the rise of a new class of political influencers who, empowered by social media’s unique incentive environment, have come to exert near-symphonic control over the fear, anger, and tribalism of large sectors of the American public. The phrase ‘political influencer’ calls to mind names like Tucker Carlson and Candace Owens, but I mean it to refer to any content creator, pundit, journalist, or even politician with an active online presence oriented around the production of political content—so perhaps hundreds of thousands of users with followings of varying sizes….

Despite accounting for only a small slice of the online population, this new influencer class is coming to dominate the market for political communication. In the process, it is transforming America’s perception of itself, which, since America is a social entity constituted in part by its self-perceptions, just amounts to saying it is transforming America.

Witkin calls the phenomenon “elite radicalization” and makes it the core of his indictment of social media platforms. Clegg not only doesn’t grapple with the research that Witkin synthesizes, he casually waves it off by arguing that Meta and other platform companies lack an “incentive” to promote extreme or hateful content. Advertisers, he writes, don’t want their products juxtaposed to such material.

Clegg’s argument is so weak as to be insulting. Advertisers, other things being equal, might prefer to hawk their wares next to nontoxic content. But advertisers, above all else, want user attention, as measured by screen time and engagement. For more than 15 years, they have demonstrated — as indicated by the astronomical revenue generated by Meta, Google’s YouTube, and more recently, TikTok — a willingness to tolerate a socially unhealthy measure of hatred as the cost of well-documented user engagement. (X’s fortunes demonstrate that advertisers have their limits. Since Elon Musk took over the platform in 2022 and turned it into a trolling cesspool, X has lost a substantial portion — although certainly not all — of its corporate ad revenue.)

Phony reformer

Clegg portrays himself as a proponent of reform by means of government regulation. Once again, his claim rings hollow.

Around the time Clegg joined Facebook in 2018, the US Congress — Republicans and Democrats alike — began a fumbling and so-far largely futile effort to rein in social media companies by means of legislation on transparency, content moderation, antitrust and other issues. The following year, Meta and other Silicon Valley companies, quietly began funding lobbying efforts under the banner of the “American Edge Project” designed to look like a grassroots movement opposed to government regulation in the name of patriotism. On its website, American Edge calls itself “a coalition dedicated to the proposition that American innovators are an essential part of US economic health, national security and individual freedoms.”

Clegg neglects to mention American Edge in his book. Instead, he makes portentous but empty declarations like, “If the masters of Silicon Valley refuse to open up, the choice will be taken out of their hands.”

He speaks favorably about a worthy but ultimately unsuccessful bill introduced in 2023 called the Platform Accountability and Transparency Act, which would have required social media companies to share more data with the public and researchers. The idea behind the bill was that by forcing the companies to disclose currently secret information about the workings of their algorithms, content moderation systems, and advertising programs, politicians and regulators would have a sound basis for proposing more substantive reforms.

Clegg ambiguously implies that he endorses this approach but fails to address an obvious question: during his time there, why didn’t Meta unilaterally make these kinds of disclosures and seek to lead (or shame) its rivals into doing the same, rather than lobby against regulation?

Instead of grappling with his former employer’s lack of self-regulation, he spews more specious verbiage: “Journalists and activists will continue to keep the pressure up on companies like Meta to be even more transparent and accountable, and rightly so,” he writes. “But any meaningful scrutiny of Meta’s policies and practices will find a company that is doing far more to keep bad content from spreading across its services than it is obliged to.” Well, yes, that’s because, in part as a result of Meta’s deceptive pressure campaign, Congress has not imposed any obligations on Meta and its rivals.

One of the steps Meta has taken to promote transparency, according to Clegg, is its occasional provision of data to outside researchers. But this is deeply misleading. Repeatedly during the time Clegg served as a top Zuckerberg lieutenant, the company restricted transparency and undermined outsiders’ ability to assess the effects of social media on users and society. Meta actually provided outside researchers with faulty data that omitted vast swaths of its user base, aggressively shut down an NYU-based project that surfaced embarrassing insights into its political advertising and misinformation policies, and killed CrowdTangle, a tool used by academics, civil society advocates, and journalists to track distortions and falsehoods on its platforms.

Harm to young users

Clegg is no better on the topic of harm to adolescent social media users. He dusts off familiar Meta talking points to make the case that the simultaneous rise in social media use and decline in teen mental health that occurred in the 2010s is a function of mere correlation, not causation. There’s a vigorous academic conflict on this topic. My former NYU colleague, the social psychologist Jonathan Haidt, leads the charge for a straightforward causal relationship in his bestselling book, "The Anxious Generation"; his detractors say the data are more ambiguous than Haidt contends.

Clegg disparages Haidt’s methodology but embraces several of his main remedies: limiting access to phones in schools, imposing more stringent age limits on social media use, and encouraging increased in-person play and other interaction among children without their screens. "We all — parents, companies, governments, and individuals — have a responsibility to the young and vulnerable in our societies,” he writes.

But if he had this responsibility, why didn’t he act on it when he was in a position to do so? Imagine the potential impact if Meta loudly endorsed those remedies and dared its corporate rivals to do the same!

Within the company, employees have worried for years about the effects of its platforms on young people. One illustration: A brief filed by the attorney general of Tennessee in litigation against Meta revealed a 2019 study the company commissioned on teenage mental health. The researchers alerted Meta that “[teens] are hooked despite how it makes them feel … Instagram is addictive, and time-spend on the platform is having a negative impact on mental health.” They added: “Instagram is the worst platform for mental health.”

A separate brief filed by the attorney general of New Mexico brought to light that in August 2021, Clegg himself forwarded an email to Zuckerberg from an employee requesting more resources to address teen mental health. Clegg annotated the communication, saying that it was “increasingly urgent” to address “concerns about the impact of our products on young people’s mental health.” He noted that the company’s “well-being work is both understaffed and fragmented.” This is not the impression one gets from Clegg’s new book. Zuckerberg, by the way, apparently did not respond to Clegg’s plea. (Haidt cites these excerpts from the attorneys general briefs in his own review of the Clegg book for The Guardian, a piece well worth reading in its entirety.)

Conclusion

Clegg addresses a number of other issues in "How to Save the Internet," including the future of artificial intelligence and the need to prevent balkanization of the internet globally. These discussions are superficial, trite, or both. He proposes, for instance, a “digital democracies alliance” to avoid fragmentation of the internet via conflicting national regulatory schemes. But he does so without explaining why companies that have tried to undermine vastly less ambitious rules would go along with such an effort or how the US, now led by a xenophobic autocrat seemingly intent on destroying existing democratic alliances, would play the role Clegg imagines in helping forge a new global collaboration.

This book fails to accomplish its stated goals. That Nick Clegg, of all people, would foist it upon the reading public marks a new low for Silicon Valley arrogance.