Summary and Review of Jacob Siegel’s The Information State

The Story of the Fusion Between Silicon Valley, the Democratic Party, and the US Security State

tl;dr: It’s a great book, but flawed. I wholeheartedly recommend buying and reading it. The Information State can be purchased here or here.

Unless otherwise noted, all quotes in this review are from The Information State and the numbers in parentheses are the page numbers the quotes are found on.

The Information State

Siegel’s thesis is concisely stated in the introduction:

What I call the information state refers to a form of government that replaces the democratic principle of consent with control. Most of us are familiar with the distinction between authoritarian regimes that impose their will by force and democracies that govern through the consent of their citizens. The information state represents a third form of political regime. It governs by controlling the codes and protocols of the digital public arena, which it uses to engineer the public’s compliance with its programs. Such a state rules by monopolizing attention. (2)

In other words, a fusion of elements of the formal state, especially the American foreign policy and national security apparatus, with the para-state, a network of ideologically driven and publicly-funded and/or privileged organizations, and major tech companies, has produced a new type of government. The information state achieves its goals by subtle manipulation rather than direct coercion or popular consent. This is made possible by modern technology, which allows for far greater central understanding of and control over people’s information environment and social networks than ever before.

This regime, the titular information state, has no constitutional or legal existence, and operates through informal chats and code rather than law, regulation, or judicial order. It is largely independent of, more powerful than, and often hostile towards the elected and formal governments:

Tech platforms like Twitter, Facebook, and Google were neither fully private enterprises nor open public squares, but something else: closely guarded artificial environments that combined politics, news, entertainment, information, and social engineering into a single medium.

Shaping that medium was the task of officials—in both the public and private sectors—who operated between the seams of their legally prescribed authorities. There was the government that consisted of its formal branches with their divided and enumerated powers. And then there was a second meta-structure of governance that used text threads, Slack chats, amorphous interagency task forces, and third-party cutouts to implement its will outside of any official charter, over and above elected officials and the formal government. (255)

Is the Information State a Useful Abstraction?

Yes. There is no better term for the consortium of formal state actors, naturally-monopolistic tech companies, and nonprofit state-funded and privileged private actors in NGOs, academia, and journalism responsible for the closure of the Internet. Treating these groups as a bloc is appropriate, as there was and is enormous overlap in personnel and concepts, constant money transfers, and explicit coordination between all three. The major alternative is Mike Benz’s or Andrew Lowenthal and Matt Taibbi’s “censorship-industrial complex,” but as Siegel correctly points out, the information state operates through much more than just censorship.

With that said, while the concept of “the information state” is useful, Siegel’s political taxonomy of authoritarian/democratic/information is not. This is because authoritarian vs democratic is itself a bad, though sadly very common, abstraction for understanding states. “Authoritarian” isn’t a useful concept at all, except as the complement of “democratic,” for which undemocratic is a better word. The reform era People’s Republic of China, the Islamic Republic of Iran, ancien régime France, England under the Whig Supremacy, Nasserist Egypt, Saudi Arabia, the Soviet Union, Nazi Germany, PRI Mexico, Tenochtitlan’s Mesoamerican hegemony, and Pinochet’s Chile are all wildly different from each other both structurally and in how they derive legitimacy.

“Democratic” is better, but what defines a democracy isn’t consent of the governed. You, personally, cannot choose not to consent to be governed even in the most democratic state. What defines a democracy is that some important officials are elected through mostly-locally-fair and reasonably competitive elections with broad suffrage, and that these elections legitimize the government. Both democratic and undemocratic states govern through a combination of coercion and consent.

Siegel’s mistaken taxonomy leads to two significant errors. First, a regime within a formally democratic state operating outside constitutional or legal restraints is not a 21st century phenomenon. In the United States, the administrative state dates back to the Progressive Era and was massively expanded, arguably enough to overshadow the constitutional government, twice, in the New Deal and in the 1960s. The national security state (sometimes called the “deep state”), of much concern to Siegel, dates back to WWII. And what I call “the para-state” (also termed “the Blob” or “the broad state”), which isn’t part of the formal government at all, has been exercising large and increasing chunks of state authority since the 1960s. Similar governing structures exist in all democracies today.

Second, a regime that operates through information control, shaping what concepts, facts, and narratives can be collectively thought about, more than coercion is not new. In describing the information state this way, Siegel has inadvertently re-invented Curtis Yarvin’s concept of “The Cathedral.” But Yarvin was writing about the Cathedral in 2008, six years before Siegel places the origins of the information state.

What is new, and why this book needed to be written, is the integration into the Cathedral/the para-state/the deep state/the administrative state of major tech platforms and modern machine-learning techniques. So rather than a wholly new regime opposed to 2013’s democracy, the information state should be seen as an extension of 2013’s regime¹ into cyberspace. I don’t hold this failure against Siegel or the book, as it’s not his fault that mainstream political science, like almost all non-technical discourse, has been frozen in amber since the 1960s.

The Free Internet and USG: 1991-2013

The Internet may have begun as a DARPA project, but as Siegel points out, the ethos of cyberspace and Silicon Valley until the 2010s was broadly libertarian, anti-bureaucratic, and suspicious of state power:

And yet the writers of the “Magna Carta” [Cyberspace and the American Dream: A Magna Carta for the Knowledge Age] were not as starry eyed as they might have appeared. The manifesto had four authors: George Gilder, George Keyworth, Esther Dyson, and Alvin Toffler. Of these, the first two worked in the administration of Republican president Ronald Reagan. They were insiders, in other words, not radicals or mystics. Their pronouncements about the overthrow of matter were not idle musings but slogans of a cyber-libertarian project to keep the government far away from the newly privatized Internet. They saw the digital revolution shaping “new codes of behavior that move each organism and institution—family, neighborhood, church group, company, government, nation—inexorably beyond standardization and centralization.” They predicted that it would spell the death of the “central institutional paradigm of modern life”: the bureaucracy. (6)

Most tech companies, subscribing to “some version of libertarianism,” had little interaction with the US government until 9/11 (128). Silicon Valley’s integration with the national security state began with helping the NSA surveil the torrents of information and communication flowing through the Internet, including the communications of Americans, with the ostensible purpose of preventing terrorist attacks.

(In defense of the NSA, while Siegel points to the surveillance itself, he never documents any instances of this surveillance being used inappropriately in this early period. It seems likely that without whistleblowers no one outside the national security state and its contacts in the tech industry would have known or even been affected. The NSA was, along with tech platforms, in the business of reading everyone’s mail, but not changing, suppressing, or censoring it.)

Silicon Valley became a more active partner of the US government with the first Obama administration. Obama was the first major Internet candidate, tech-savvy and using the Internet to run rings around newspaper/TV gatekeepers, in the 2008 Democratic presidential primary. In his first term, he built an alliance “between his party and the tech industry that fundamentally changed what the Internet was” (140).

Siegel uses Google to illustrate this transition. Per Siegel, “Google originally conceived of itself as a digital Switzerland, a high-trust society where citizens relied on a neutral civic infrastructure,” reflecting both Silicon Valley’s libertarian culture and the knowledge that “being seen as fair and neutral bolstered a company’s esteem in the eyes of users” and taking sides in domestic politics would alienate half the country (140). This changed with the Obama administration, which Google was closely involved with from the beginning. In Siegel’s words, “Google increasingly operated like a branch of the US government. On everything from personnel to policy, the platform and the state appeared to be merging” (159). Google met, on average, more than once per week with White House officials 2009-2015, with “252 cases of staffers moving between jobs at the White House and the tech giant” (159).

Silicon Valley acting as an active agent of the US national security state began with foreign policy. In 2009, Obama’s State Department, under Hillary Clinton, devised the concept of “digital diplomacy,” using “nominally private tech platforms as arms of state power” (148). By working through private companies, the Obama administration could circumvent public and legal deliberation:

Instead of working through Congress or executive actions, the Obama administration could circumvent public review by getting the tech companies to implement its foreign-policy aims through their corporate policies, “terms of service,” and other changes to the digital code. Regulating how much access foreign countries had to digital networks became an important form of leverage in US foreign policy. Washington pressured the tech companies, which pressured the rest of the world. The public had no role to play. (148)

Crucially, however, this alliance between the national security state, the Democratic Party, and Silicon Valley was in this era on the side of Internet freedom. Structurally, it was quite different from the pre-9/11 cyber-libertarian ethos and even tech’s passive cooperation with government surveillance 2001-2008, but the actual effect was similar. This is because the Democratic Party and national security state agreed that Internet freedom served US goals by enabling preference cascades that would destroy unpopular and ossified authoritarian hegemonies:

In 2010, the Clinton State Department unveiled its signature “internet freedom” agenda. It would be one of the last gasps of 1990s-era techno-optimism. In a speech that year at the Newseum in Washington, DC, Clinton warned about the spread of digital censorship in authoritarian regimes. “A new information curtain is descending across much of the world,” she told the audience. “And beyond this partition, viral videos and blog posts are becoming the samizdat of our day.” Within a few years, Clinton would completely reverse her stance. Once Donald Trump rose to national power, she, along with every other senior official in the Democratic Party, would attack, as the single most dire threat to American democracy, the Internet openness she had once lauded. (149)

And this pro-Internet freedom agenda was not just talk:

On the ground, the Internet freedom and digital diplomacy initiatives delivered tens of millions of dollars in funding to US-aligned NGOs in eastern Europe and the Middle East. Government grants went to NGOs whose charters were to bring millions more people online. The State Department “encourage[d] activists around the world to use high-tech tools.” At the same time that it was enlisting activists onto the US payroll, the agency was “freeing up export rules so social-media firms such as Facebook, YouTube and Twitter can access markets in places that used to be strictly off-limits, including Iran, Cuba and Sudan.” (149)

The Democratic Party/Internet freedom/national security alliance culminated in 2011 with the Arab Spring, which had the open and direct support of the Obama Administration (150-151). Many of the organizers were cadres trained in previous years by US-funded NGOs in using digital technologies (incidentally, Hosni Mubarak, the Egyptian leader deposed by the Arab Spring, had been a close US ally for decades).

But the fragility of this alliance can be seen in the way Clinton’s senior adviser for innovation at the State Department, Alec Ross, endorsed it: he “hailed social media’s role in the Arab uprisings by comparing it to ‘the Che Guevara of the 21st century’” (151). Che Guevara is not known for his libertarianism or love of freedom; a perspicacious observer might have been able to deduce from that statement that the Democratic Party and national security state’s endorsement of Internet freedom was contingent on Internet freedom serving the Democratic Party’s ideological goals—and these goals were not limited to overthrowing US allies abroad.

2014: Hybrid Warfare

In Siegel’s telling, the US national security apparatus first turned against the free Internet in 2014, in response to three events: the Euromaidan movement in Ukraine, the Russian invasion of Crimea, and ISIS’s capture of the Iraqi city of Mosul (170). Russia, Russia, and ISIS, respectively, were seen to have used “information dominance” on social media to confuse and demoralize opponents rather than triumphing with conventional military might (175). This was taken to imply that Internet freedom, rather than liberal democracy’s spear, was actually an Achilles heel that its (still foreign, at this time) enemies could exploit, and one that no amount of conventional military force could defeat.

This was mostly wrong. The pro-Russian faction lost in Ukraine, Russian soldiers were much more important to conquering Crimea than Russian social media propaganda, and precisely-placed high explosives turned out to be much more important to defeating ISIS than control over social media. But it was useful for national security officials, who could then explain their failures “in terms of grand narratives about civilizational values and dark Internet magic, rather than as a consequence of mistaken assumptions, missed opportunities, and policy failures” (175). Even better, “it argued for granting more authority over political matters to unelected security officials who policed the information space” (175).

Still, in 2014 the national security state’s abandonment of Internet freedom, and embrace of online propaganda, was still targeted at foreign adversaries. This was the age of the “Information Coordination Cell” putting out what it called “terror facts” (a precursor of later “fact-checking”) to debunk ISIS propaganda, not mass censorship of Americans on social media (177).

“Whole of Society”

The title of Chapter 9 of the book is titled “Whole of Society.” This bland-sounding piece of NGO-speak is a euphemism for the totalitarian idea that every powerful organization or institution ought to serve the dictates of the state. An elementary school cannot just teach students reading and math. A shoe factory cannot just make shoes. A bank cannot just provide financial services. Passive index asset managers controlling a significant fraction of private capital on behalf of millions of small shareholders cannot passively invest. Instead, all of these organizations must put their weight behind whatever the ideological project of the day is. “Whole of society” represents the destruction of any remaining divisions between the public and private sectors, the quenching of the last embers of the classically-liberal order America was founded on.

As a concept, “whole of society” originated in the world of international NGOs, migrated to counterinsurgency manuals during the wars in Iraq and Afghanistan, and was imported to American domestic politics by the Obama administration (155). By weaponizing non-formal-state chokepoints in American society, the Obama administration learned to fabricate apparently-organic consensus on contested and controversial issues, and then enforce this false consensus on the rest of society unbound by law or formal procedure.

In the “whole of” sense, society now described a select set of stakeholder organizations brought together to execute state policy. The term became a euphemism for powerful institutions, including banks, universities, and social media platforms that did not represent the public but controlled its access to resources like financial services, social connection, and professional accreditation. The new society was shallower, less rooted in the realities of human relationships and the physical environment, more prone to sudden fluctuations. At the same time, however, it had significant advantages as a governing construct. With fewer points of friction, it could be more efficient and better able to implement sweeping, system-wide changes, like a phone automatically updating its operating code. (155)

Anyone who lived through the Great Awokening can think of examples. The only analogues in pre-Obama American history were during total wars.

Mass Censorship on Social Media

With the Silicon Valley/Democratic Party/national security state alliance sealed in 2008 and turned against Internet freedom in 2014, and the techniques of whole-of-society mobilization of all significant institutions, private or not, for political ends adapted to domestic American politics by the Obama administration, the stage was set for mass censorship and the creation of the information state proper. In Siegel’s telling, the Democratic Party began moving in this direction in 2015, but was initially resisted by still-residually-libertarian tech companies, especially Mark Zuckerberg. It was Trump’s election in 2016 that triggered the true crackdown:

For Clinton and Obama in particular, the 2016 election result pointed to a profound betrayal. Through the 2000s and early 2010s, they had both evangelized for Silicon Valley while turning a blind eye to its flagrantly monopolistic business practices and ties to China. In return, they had expected the tech platforms to censor pro-Trump messaging under the guise of purging fake news. Instead, the social media companies partially resisted that pressure in 2015 and 2016. While agreeing to stricter moderation of content, Facebook and others insisted that they were neutral platforms and should not be policing protected speech. But like the ideological activists who dismissed free speech as a tool of white supremacy, leading Democrats attacked the platforms’ neutrality as a terrible crime. (207)

This crackdown was legitimized by portraying Trump’s election as the result of foreign, and specifically Russian, disinformation campaigns on social media. Siegel documents² how unanimously the security state endorsed this (false) narrative, and how studded both tech companies and the para-state³ became with ex-(or “ex”)-national security state officials. FBI, CIA, DoD, NSA, State Department, DHS, and other intelligence alums entered the private and para-state sectors and set about working hand in glove with their old organizations to purge “disinformation” and “foreign influence” from cyberspace.

In fall 2017, the FBI opened its Foreign Influence Task Force “to identify and counteract malign foreign influence operations” through “strategic engagement with US technology companies.” The office included representatives from the FBI, the CIA, the Department of Homeland Security, and the Office of the Director of National Intelligence. A dedicated Slack channel allowed security officials to communicate directly with social media executives about accounts they wanted to flag or take down. Censorship became a routine and informal agreement between the various parties who had been brought together in the fight against disinformation.

Fact-checkers, content moderators, nonprofit anti–hate speech representatives from groups like the Anti-Defamation League, counterterrorism veterans, trust and safety officials, countering violent extremism experts, social scientists, political operatives, FBI agents, millennial journalists, and CIA officers all rubbed shoulders on the counter-disinformation party bus housed inside the social media companies. (211)

The Democratic Party became the elected arm of the information state, adding the power of state governments and parts of Congress, and especially threats of anti-monopoly regulation⁴ if tech companies didn’t play ball. This has been documented elsewhere; by 2018 the second most common background for Democratic candidates, after “lawyer” was “national security official.” As with the initial alliance between the Party and Silicon Valley, Obama was personally important, signing an executive order creating a propaganda/censorship/narrative control department within the State Department on his way out of office.

In his last days in office, President Barack Obama set the country on a new course. On December 23, 2016, he signed into law the Countering Foreign Propaganda and Disinformation Act. Against the backdrop of the extraordinary allegations made about Trump’s collusion with Russia, it used the language of defending the homeland to launch an open-ended information war whose primary target would be the American public. The act directed the State Department to expand the mission of the Global Engagement Center, an office established earlier that year by executive order whose mission was to counter the effects of foreign propaganda and disinformation. Under its new charter, the GEC would lead a whole-of-society campaign to align the most powerful actors in the public and private sectors with the policies of the ruling party. (201)

Siegel has many examples of censorship itself. Some of these were familiar to me before reading the book, such as Hamilton 68 smearing hundreds of Anglosphere conservatives as Russian bots to then portray their concerns as downstream of Russian influence (212-213). Other were not; I had never heard of the New Knowledge project that created fake conservative Facebook profiles supposedly backing Republican Senate candidate Roy Moore and then had the chutzpah to attribute these fake accounts (which they created) to a pro-Moore Russian disinformation campaign (221-222). Some of the most directly electorally-relevant were the public-private “Election Integrity Project” (225), and above all the censorship of the New York Post’s (entirely true) reporting on Hunter Biden’s laptop as a foreign disinformation operation (226-227), demanded by dozens of US national security officials. I am tempted to excerpt them all, but at that point you may as well just read the book.

Siegel correctly observes that the censorship apparatus went into overdrive during the lockdowns, banning disagreement with public health authorities even as said authorities changed their line daily (233), including explicitly true information that people might take the “wrong” way (234). This culminated with the deplatforming of president Donald Trump in January 2021 by a Silicon Valley cartel, which showed just how powerful and independent of the elected government the information state was (236).

He also points out how the information state actually increased in power initially under Biden, adding the presidency to the long list of institutions and power centers involved. The Biden Administration, though perhaps not Biden himself due to senility, was directly involved in expanding and intensifying the censorship-industrial complex in 2021 and 2022. “Conflating terrorism, political extremism, and information, the administration’s policies blurred the lines between ideas, actions, speech, and violence, while extending its policing power over all of them” (239).

Fact-Checking

Siegel covers the rise of the “fact-checking” complex, which determined which narratives, facts, and talking points were regime approved and which should be suppressed. This began in 2015 with the foundation of the IFCN, the first and at the time only major credentialing body for fact-checking organizations, which was launched as “a division of the Poynter Institute, a media nonprofit closely aligned with the Democratic Party” (209). The IFCN, not being part of the formal government, lacked both official power and congressional or Constitutional constraints, and acquired authority with the threat of anti-trust action and NGO lawsuits.

Unconstrained by congressional oversight or the Constitution, the IFCN did not require any statutory authority to pressure a company like Facebook into accepting its edicts. Rather, the large tech platforms sought to inoculate themselves from government regulatory pressure that would affect their profits and from the threat of lawsuits from the NGO sector by accepting the authority of the fact-checkers. It was a concession to the not-so-subtle threat that if the tech giants didn’t start censoring themselves, they might get their windows, or their monopolies, broken by the state. (209)

All of this rested on the lie that apolitical “fact-checker” was a real profession in the American media landscape circa 2015 which existed and could be deferred to, which was not true:

In its letter to Zuckerberg, the IFCN sold the idea that an army of professional fact-checkers stood ready to enter the fray and battle disinformation on behalf of the public good. The pitch rested on the false premise that a recognized hierarchy of “prominent fact-checkers” was a real thing. No such profession existed. In the American media landscape, fact-checking had long been an apprentice-level job mostly reserved for recent college graduates who ensured that busy reporters correctly reported the date of Moldova’s first democratic election after the fall of the Soviet Union. What the IFCN offered was not fact-checkers in the traditional sense, but a cadre of compliance officers who would scour the Internet, flagging anything that threatened the interests of the ruling party. (209)

Beyond Censorship

What differentiates Siegel’s concept of the information state from terms such as “the censorship-industrial complex” is that the information state is not limited to censorship.

Narrative Enforcement

One technique of the information state is the summoning into existence of regiments of credentialed fake experts (meaning lacking real expertise). These “experts” are employed in many nominally-independent organizations, which can then be cited by sympathetic journalists and Wikipedia, such that there appears to be a broad-based consensus that a course of action is correct. Siegel uses the JCPOA agreement with Iran in 2015 as an early example:

Samuels, the journalist who elicited Rhodes’s frank admissions, provided an early description of the newly developing media complex that clustered around institutions funded by the Democratic Party. “Legions of arms-control experts began popping up at think tanks and on social media, and then became key sources for hundreds of often-clueless reporters.” When asked about the “onslaught of freshly minted experts cheerleading for the deal,” Rhodes explained how the White House had manufactured a consensus: “We created an echo chamber.” The legions of experts were actually apparatchiks. (182)

But of course these “experts” are not real experts, only imitating the trappings of expertise, and the institutions that employ them are not truly epistemologically independent, but part of a dense network linked by funding, personnel transfers, and milieu.

What Siegel has rediscovered here is, once again, Curtis Yarvin’s Cathedral. The sheer weight of hundreds of prestigious and authoritative institutions and thousands of credentialed professionals all agreeing on one narrative is very difficult for any would-be dissident to oppose. Can Harvard and the Washington Post and Politifact and the State Department and Amnesty International and the New York Times and the WHO and the European Union (and…) all really be lying?

The way out is to understand that all of these institutions and credentialed professionals are not genuinely independent of each other. They are like many megaphones attached to one microphone, all repeating a message. This message is not necessarily wrong, but 1000 connected megaphones provide no more evidence than one. Similarly, institutional consensus provides no more evidence than a single one of these institutions would by itself.

The JCPOA is a minor case, but a much more important instance of this technique was the “Russiagate” narrative, that Donald Trump was (de facto) a Russian agent who reached the Oval Office as part of a devious Putinist plot to subvert Our Democracy.

Russiagate

I never paid much attention to Russiagate back in 2016-2018, because it seemed like an obviously absurd astroturfed conspiracy theory constructed for purely cynical partisan purposes, not worth learning the details of. At the time, I had enough faith in the US political system to believe that since there was nothing serious to it, it would eventually dry up and blow away, without significant long term effect.

I was right that the Russiagate narrative was nonsense and would eventually dissipate. But I was very, very wrong about the lack of effect. At the time, I did not appreciate the power that lies, even lies so obvious that any half-way sapient independent observer can clock them in a minute, can have for coordinating political coalitions⁵.

The important thing to understand about Russiagate is who and how many “experts” were involved, including a significant chunk of the US national security apparatus, academia, and media:

One of the most disorienting aspects of the conspiratorial mania that overtook America’s elites in response to the rise of Donald Trump was the sheer scale of expert consensus behind views that were, on their merits, utterly deranged. What an ordinary person saw in 2016 was the country’s most venerated institutions all promoting the same claims about a Russian takeover of the American political system. Any given charge about Trump’s ties to the Kremlin might fall apart under scrutiny, but there were so many, coming from seemingly authoritative sources, that their totality seemed to outweigh their individual merits. The alternative—that it might all be so much propaganda—was difficult to face. It meant that the legions of Harvard professors, senators, senior national security officials, and respected journalists touting Trump’s sinister connections to Vladimir Putin had allowed themselves to become credulous bullhorns for a cynical and destructive information operation. (187)

As loath as I am to use the term, Russiagate was an extraordinarily successful disinformation operation. By redefining American conservatism as a threat to democracy downstream of foreign subversion, the information state was able to circumvent the United States’ institutional, legal, and cultural checks against mass domestic censorship and redirect the conceptual and technical infrastructure built for foreign “information operations” domestically. This legitimized the closure of the Internet to traditionally-conservative—or at least RedGov—power centers that might otherwise have opposed transparent ideological censorship. By the time it all fell apart in 2019, the censorship-industrial complex was powerful and entrenched enough to do away with the national security justification.

Debanking

When censorship and narrative enforcement fails, the information state still has other tools with which to obtain compliance without cracking skulls. The most pernicious of these is debanking, but almost any system dependent on network effects or with high fixed costs can be (and was, and is) similarly weaponized.

In Canada, after a group of truckers started a mass protest of their country’s vaccine mandates in early 2022, Prime Minister Justin Trudeau declared a state of emergency. The Canadian government directed private banks to freeze the accounts of at least 257 individuals who donated to the trucker convoy, denying them access to their own money. Debanking, as the practice was called, spread rapidly throughout other countries. In the US, numerous January 6 defendants who were sentenced for misdemeanor charges related to the Capitol riot reported being debanked and banned by professional services like LinkedIn. Separation between public and private spheres had for centuries constituted the essence of liberal society. As the walls between them collapsed, state-directed informational power rushed across the breach.

Debanking was one instrument in the larger arsenal. The whole-of-society framework had erected a 360-degree police force made up of the companies a person did business with, the civic organizations that were supposed to make up their communal safety net, even their neighbors. A small group of powerful people who controlled key chokepoints in the digital network used public-private partnerships to deny anyone access to money, credit, social platforms, and other public accommodations. (242)

What Siegel fails to point out in The Information State is that while denial of access to financial services of citizens (as opposed to foreigners who could be subject to sanctions) without some judicial process by formal states was pioneered by Canada in 2022, debanking of dissidents began several years earlier, coordinated by “civil society” para-state actors like the SPLC.

The Failure of the Information State

In Siegel’s telling, the information state imploded with the Biden administration. The censors, coming as they did from fields with a history of failure like counterterrorism, journalism, academia, and epidemiology, did not accomplish what they set out to. The information state did not “deradicalize” Trump supporters, it did not convince the general populace to obey the whims of public health during the lockdowns, it did not keep Trump out of office for a second time, and it did not rebuild trust in the affiliated institutions of academia, journalism, and the elected Democratic Party (249).

Siegel charts the collapse of the information state as follows: first, Elon Musk’s 2022 purchase of Twitter put a hole in the previously-monolithic panopticon, triggering a preference cascade among other tech leaders and revealing many of the mechanics of the information state with the Twitter Files (252-255). The 2023 court case Missouri v. Biden, which accused the Biden administration (correctly) of conspiring to mass censor American citizens, brought much of the previously-hidden and opaque censorship system into the public eye (256). With the political winds shifting, Mark Zuckerberg publicly regretted acquiescing to White House pressure to censor Facebook (256). Biden’s senility, which had been carefully covered up by journalists and fact-checkers for four years, was made so obvious it could not be denied any longer in his presidential debate with Donald Trump, discrediting the official truth portion of the information state (257). Finally, Trump himself burst the bounds of the narrative control apparatus after the failed assassination attempt in Butler and entered the Oval Office for a second time in triumph (258). Thanks to the relatively free discourse on X and the tireless efforts of anti-censorship activists, Trump II understood the information state from the beginning, and immediately began purging it from the formal government (259).

This is a triumphant narrative: the dragon slain and democracy restored. It sits uneasily with Siegel’s core thesis that the information state represents a new kind of political regime on par with democracy or authoritarianism. Regimes worthy of the title rarely collapse so quickly and easily. And it is wrong.

The censors have been largely purged from the formal US government. But the formal US government was only ever a small part of the apparatus. The legions of disinformation experts, moderators, fact-checkers, journalists, and intelligence agents embedded in the para-statal quasi-public sector, the major tech platforms themselves, and above all in many foreign governments and institutions are all still there. Using sympathetic governments in the EU, Canada, and the UK to continue the closure of the Internet by putting pressure on tech companies is the explicit strategy of many of them⁶.

At the moment, they are restraining themselves from too-overtly targeting the mainstream Republican Party and associated ideas and institutions for fear of retaliation from the Trump administration, which is a good thing. But that only lasts as long as Trump, or his heir, is in office. And while the GOP proper, along with a cluster of contrarian alt-center high profile podcasters, is mostly safe at the moment, the galaxy of heterodox and dissident thought snuffed out by the closure of the Internet remains suppressed, especially off X and Substack. And even on those platforms, many anons have been doxxed by the professional enforcement apparatus of the information state, with corresponding chilling effects.

What I Would Have Liked

I was recommended The Information State as a book length treatment of the closure of the Internet, the conversion of the Internet from an almost entirely free and open space to a tightly controlled and censored playpen in the late 2010s. In my view, the closure of the Internet is the single most important social phenomenon of the late 2010s, and all coverage of it that I’m aware of is scattered and fragmentary. The information state is closely related to the closure of the Internet, but The Information State did not quite fill the niche I wanted.

The book focuses on the GWOT, the US national security state, and the Obama/Biden administrations. There is very little attention paid to non-US state actors, like the EU, Germany, Britain, or Australia. This would be a serious oversight for a book on the closure of the Internet; Germany was threatening Facebook and Twitter over anti-immigration speech in 2015, and the EU imposed a “voluntary” commitment to censor “hate speech” on most major platforms in May 2016. This was before Trump’s election and had nothing to do with ISIS or Russian interference.

(In fact, the left wing of the German establishment and the EU were broadly pro-Russian before the 2022 invasion of Ukraine, hence amusing episodes such as the Financial Times arguing that Polish opposition to Nordstream II, a flagship prestige and economic project for the Russian state, was actually pro-Russian because it undermined Germany and European unity.)

Nor is there much time spent on the actual platforms censored, when and how the platforms cracked down, non-state actors like the SPLC except when these were tied to the security services and foreign policy apparatus, or the effects on would-be dissidents or heterodox discourse. Most of the concrete instances of censorship focus on mainstream Republican views and opposition to the lockdowns. These of course should not be censored, but they’re not where the closure of the Internet began. The focus is on GOP-aligned institutions and individuals, and not on the far larger, more diverse, and more original set of people outside of knowledge producing and disseminating institutions entirely. As a work of political science explaining the structure of present-day Western government, parts 2 & 3 are fantastic, but there’s not much in the book on the effects from the perspective of a random user/citizen/poster or on the benefits of open discourse and freedom of speech more generally.

If I were a political operative, I would take the same state-focused tack as Siegel, as spy chiefs censoring American political speech is unconstitutional, extremely unpopular, and fairly addressable through political action by elected officials once the thick layer of euphemisms (“disinformation,” “online safety,” etc) is scraped off. But my chief concern is not on who exactly is responsible for the censorship, but that it’s being done at all. This is particularly true post-2024, as Trump II has dismantled most of the formal US state-linked components of the information state (for which he and his administration have my gratitude), leaving the locus of the problem abroad and in the private and para-state sectors. The Information State is also mostly a compilation of secondary sources, which is absolutely worth doing, but I was hoping for some original interviews or archival research.

These are not criticisms of The Information State. The story of the nexus between tech companies, NGOs, US security services, mass censorship, and the Democratic Party is extremely important. So is the story of how the infowar practices developed to fight Russia and ISIS were turned against domestic opponents. And even though it is not intended to be an overview of the closure of the Internet, it is still the best and most comprehensive single work I am aware of on the subject. The world is better for this book having been written and would be improved further if its findings and vocabulary became common knowledge. Again, go buy it.

Appendix: Dislikes

When I was compiling excerpts for this review, I categorized passages as good (useful, informative, insightful) or bad (false, misleading, pointless). Aside from the introduction, every single passage I classified as good came from after page 120, and every passage I classified as bad came from before 120. I have put these bad parts in an appendix because they do not really impinge on the book’s main thesis or evidence, but they cover enough of the book to be worth addressing.

Critiques Frozen in 1960

The core issue is that the book combines a compelling thesis, strong narrative, and reams of evidence about the 21st century information state with a mediocre and often false historical narrative of the history of information and of 20th century America. The trouble is that Siegel is writing within the New Left intellectual tradition. The New Left, and many other related radical thinkers in the postwar era, were generally hostile towards technological, industrial society, modernity, and the notion of material progress, in part because that era’s capitalism was so effective at delivering it. Siegel inherits that hostility. This is not necessarily a problem; there are many valid critiques of techno-industrial progress and modernity. The problem is that he also inherits the view that these things characterize contemporary America, which was true back when these ideas were formulated but not any more.

This is not unique to him or to this subject. Western discourse on most issues of consequence has been frozen in amber since the 1960s, because New Leftists successfully politically captured our sense-producing institutions (social science and humanities academia and through it K-12 schooling, media, entertainment) and have since gradually frozen out heterodox thinkers, a process just about complete today. This is why we, meaning academia, entertainment media, and journalism, still speak of long-vanished white supremacy, cut-throat capitalism, patriarchy, Christian dominance, colonialism, and other features of a world that hasn’t existed for generations, as though they’re live issues.

Whether these criticisms were right or wrong back in the 1960s and 1970s when they were being made, they are long outdated today. Just as we do not live in a society that tells girls to be housewives, but in one that tells girls that society tells them to be housewives, we do not live in a society that worships techno-industrial modernity, technocracy, and meritocracy, but one that constantly complains about how society worships techno-industrial modernity, technocracy, and meritocracy.

(Ted Kaczynski, who would have been a worthy addition to Siegel’s bibliography alongside Jacques Ellul, noticed this problem, that would-be dissidents often parrot the System’s own “critiques,” back in Industrial Society and its Future.)

This blindness leads to some howlers early on. For example, on page 15, Siegel writes: “For thousands of years, human civilizations did something unimaginable by modern standards: They set rules to limit the power of their technologies.” But modern civilizations set rules to limit the powers of our technologies all the time! We have many thousands of rules limiting the powers of our technologies, far more than any ancient civilization.

Similarly, on page 22, Siegel describes early 20th century American progressivism like so, with the implication that the American worship of technological progress described 21st century America, the main subject of his narrative:

Conversely, great engineering feats illustrated the power that human ingenuity could achieve when technology liberated itself from ethical and natural limits. The two elements, religious and technological, appeared to be opposed but had a way of merging in the American worship of progress.

A distinctively American technocratic ideology coalesced toward the end of the nineteenth century. In response to the upheavals of industrial society, the technocrats sought “to invent industrial-strength tools of social control,” writes the historian John M. Jordan. Through concepts like scientific management, they tried to reorient the economy away from self-interest and profit by placing it in the charge of technical experts.

This is a reasonably accurate description of America between WWI and 1973. But Americans have not worshipped progress since then.

Source.

It’s helpful to compare 21st century America with its pre-1970 counterpart, or the contemporaneous and similar (in this respect) Soviet Union. Both the US and USSR had serious plans to redirect entire rivers flowing into the Arctic to arid southern regions, the American Southwest and Central Asia, reshaping hydrology on a continental scale. And both actually did build continental-scale industrial infrastructure, most notably the interstate highway system, and electrical capacity in just a few years. Today, the US can’t even build a few miles of high-speed rail or reclaim more than a tiny fraction of the land it did in the 1960s.

The only 21st century state with the sort of industrial, material, and technocratic ambition and capability that characterized mid-20th century America is the People’s Republic of China, and much of the rest of Asia is far more culturally, technologically, and industrially progressive than the United States is today.

The New Left hatred of modernity, techno-industrial progress, technocracy, and meritocracy won almost 60 years ago, like its attitudes towards race, patriarchy, sex norms, and most other issues of consequence, and now dominates both discourse and American and Western civilization. It’s not 1960 any more.

Factual Errors

Siegel makes some other significant factual errors in his historical narrative. For example, he writes that:

In practice, the first phase of post-industrialism involved widespread deregulation of key industries and the first moves to offshore America’s manufacturing base. The very conditions that had produced the broad prosperity of the American middle class in the three decades after World War Two—the labor boards and legal regulations protecting family wages, pension plans, and vacation allotments—were taken to be the sources of America’s economic decline and systematically dismantled. (103)

There are several false statements here. First, labor boards and legal regulations protecting family wages, pension plans, and vacation allotments were not the conditions that produced the broad prosperity of the American middle class in the postwar era. Instead, post-war Americans rapidly grew more productive, as scientific, technological, and organizational improvements developed in the first half of the 20th century were finally, after the calamities of the World Wars and Great Depression, diffused through the civilian economy in both the United States and the entire Western world. This combined with (1) very high moral standards across all classes, meaning work ethic, discipline, and honesty, (2) a cultural and legal environment favorable to marriage, and (3) a favorable demographic situation, with the lowest immigration in over a century, to produce the broad-based post-war economic miracle. The labor boards, pension plans, and legal regulations protecting family wages were made possible by the prosperity, not the other way around. All of these pillars except the technological ones were undermined in the 1960s⁷.

Second, the US did not begin significant deindustrialization until the early 2000s when China was granted most-favored-nation status. Some US industries suffered from competition with German and especially Japanese industry in the 1970s and 1980s, and this actually was, not simply “taken to be,” partly downstream of US unions and regulations. But this did not lead to deindustrialization. Instead, American manufacturing began its long exodus from the high-union-and-regulation North to the Sunbelt. The number of American industrial workers, in my view a better proxy for industrialism than value-added calculations, stayed roughly constant between 1980 and 2000.

(The apparent decline seen in time series is an artifact of manufacturing businesses going from keeping services such as janitorial work in-house to contracting other firms for the same work. In time series, this looks like manufacturing workers becoming service workers, but about the same number of people were still doing about the same jobs until the early 2000s.)

The purpose, which is not necessarily Siegel’s but that of the tradition he’s drawing on, of mis-dating American deindustrialization by 20 years is to wrongly indict Reagan and “neoliberalism” and promote more state control over the economy as a remedy. Something similar occurs in discourse on British economic history, with the decline of British heavy industry mis-dated to the Thatcher era when it actually began around 2002.

The other major error I noticed was Siegel’s claim on page 92 that “the FBI sent an anonymous letter to Martin Luther King Jr., threatening to reveal his extramarital affairs if he did not kill himself first.” The purpose of this claim is to indict US security services. But it’s extremely misleading, because the FBI was generally pro-Martin Luther King Jr and pro-Civil Rights. During the 1960s, the FBI knew about, and repeatedly acted to cover up, King’s sexual misbehavior (and arguably crimes) and extensive ties with communism, the revelation of either of which would have devastated his moral standing. The FBI also sent agents to guarantee his protection in summer of 1964, and destroyed extra-legal southern resistance to Civil Rights by extensively infiltrating the Ku Klux Klan (Orton 2024). The FBI’s attitude towards King was one of trying to protect his reputation and immunize him from communist blackmail, not trying to destroy him. The letter is real and from an FBI agent, but likely one acting alone on his own accord without the knowledge of the Bureau.

The less-misleading framing of the story would also have served Siegel’s narrative, since politicized intelligence agents using their skills and resources to act of their own accord are themselves a huge problem, arguably larger than the politicization of the organizations themselves. A common technique is leaking private information to aligned media outlets or NGOs to be used to dox dissidents (“parallel construction”) so as to escape bureaucratic regulations and legal restraints on the domestic powers of the security services.

The source of these errors is the same as the source of his mistaken views on America’s attitude towards technology and modernity: both are very common in New Leftist (“liberal,” which I put in scare quotes to distinguish from classical liberalism) historiography, and this is Siegel’s intellectual tradition.

Boilerplate

Chapters 2 and 3, on the nature of progress and information, are mostly not even wrong. They’re not exactly factually wrong, but they’re not right either. They’re just there, taking up space without insight or relevance. The book does not require or benefit from a generic history of computing or pro forma references to Bentham’s panopticon. The digression on the use of statistics in the Vietnam War in chapter 4 is similarly pointless. That the US military uses mathematics and computing to achieve its goals, and that this does not always work, is trivial and obvious. The component of this important to the thesis, online information warfare, is covered more than adequately by post-Cold War sections.

This book would have been considerably improved if the five chapters comprising part 1 were excised, and the book simply began its narrative in the 1990s. Part 1 does not greatly mar the book, since it can simply be ignored without loss, but were I not already familiar with the subject matter of parts 2 and 3 I would have been worried that I was suffering Gell-Mann amnesia while reading them.

1. The extent to which this regime should be called “democracy” is debatable.

2. For someone looking for online reading, much of this can be found at “Report on the Censorship-Industrial Complex: The Top 50 Organizations to Know,” “An Insider’s Guide to Anti-Disinformation,” and scattered through the reports of the Foundation for Freedom Online.

3. “Non- or bi-partisan” “independent” organizations focused on “disinformation” or “hate speech” or “foreign influence operations.” It is impossible to write about this subject without copious scare quotes because of how Orwellian it is.

4. The true argument that, as natural monopolies, major tech companies act as a form of government can be used to either argue against mass “private” censorship or used as a threat to force mass “private” censorship. There was a flood of papers, articles, and books arguing for the latter 2016-2019.

5. Which should be distinguished from the utility of lies for propaganda or persuasion. The purpose of lies for propaganda or persuasion are to give the audience a false picture of the world. They are intended to be believed. On the other hand, coordinating lies are not so much intended to actually be believed as to allow members of a coalition to publicly pretend to believe them to justify accomplishing some other goal.

6. The Foundation for Freedom Online writes regular reports on the current plans and maneuvers of the censorship-industrial complex. I would recommend reading through some.

7. See Charles Murray’s Losing Ground and Coming Apart for some book length treatments, or my “The Baby Boom” for a much more cursory overview of this era.