United States Treasury Secretary Scott Bessent speaks to members of the media outside the White House in Washington, Thursday, Aug. 20, 2026. (AP Photo/Jacquelyn Martin)

What the Bond Market Disruption Means for Rebuilding American Tech Policy

Indian Prime Minister Narendra Modi addresses the nation from the rampart of the 17th century Mughal-era Red Fort monument during the country's Independence Day celebrations in New Delhi, India, Saturday, Aug. 15, 2026. (AP Photo)

Technocracy Won't Debug India's Education Crisis

House Judiciary Committee Chairman Jim Jordan, R-Ohio, conducts the Select Subcommittee on the Weaponization of the Federal Government in the Rayburn Building on Thursday, July 20, 2023. 

How the 'Censorship-Industrial Complex' Narrative Went From Fringe to Trump Policy

Abstract image of a whistleblower amid a see of faces.

Congress Can Get AI Whistleblower Protection Right On Its First Try

Nine small images with schematic representations of differently shaped neural networks, a human hand making a different gesture is placed behind each network.

AI Audits Need a Power Test, Not Just a Fairness Score

The Mathare River is one of the rivers that flows into Nairobi River. Photo by Queen Asali/Wiki For Sustainable Futures 2026/CC by 4.0

The Mathare River is one of the rivers that flows into Nairobi River. Photo by 

Taxing Tech in Age of AI Could Fund Human Rights

Composite. Left, Supreme Court Justice Samuel Alito attends a meeting in Rome, Sept. 20, 2025. (AP Photo/Andrew Medichini) Right, Meta CEO Mark Zuckerberg, CEO of Meta, is seen in the US Capitol on Thursday, March 26, 2026. (Tom Williams/CQ Roll Call via AP Images)

How the Supreme Court and Big Tech Both Weaponize 'Race-Neutrality'

Children’s Digital Rights Need More Than a False Promise of Safety

In an aerial view, the Meta El Paso Data Center is seen on August 13, 2026 in El Paso, Texas. Meta continues construction on its more than $10 billion data center, occupying approximately 1,000 acres, and has expanded the project to 1 gigawatt of computing capacity, allowing the facility to support a growing range of AI workloads and the training and deployment of AI models and technologies. (Photo by Brandon Bell/Getty Images)

Zuckerberg’s AI Manifesto Lost it at Compute

Flock Safety camera on Haverhill Drive near Phillips Avenue in Toledo, Ohio. (Photo by Stephen Zenner / SOPA Images/Sipa USA)(Sipa via AP Images)

Unless Flock is Reined In, Mass Video Surveillance Will Grow Even More Pervasive

Protesters rally against data centers in Hillsboro, Ore., on May 16, 2026.  (Photo by Alex Milan Tracy/Sipa via AP Images)

How to Use Your Community’s Data Center Moratorium

Chinese President Xi Jinping, center, waves as he arrives at the opening ceremony for the World AI Conference in Shanghai, Friday, July 17, 2026. (AP Photo/Ng Han Guan, Pool)

Will China Crack Down on Open-Weight Models?

Brazilian presidential candidate Ronaldo Caiado, former governor of Goias state, presents his public security policy in Sao Paulo, Brazil August 10, 2026. REUTERS/Jean Carniel

How a Brazilian Governor Turned Rare Earths Into a Campaign—and Washington Into a Partner

CEO of Meta Platforms Mark Zuckerberg attends the Allen & Company Sun Valley Conference at the Sun Valley Lodge on July 9, 2026 in Sun Valley, Idaho. (Photo by Kevin Dietsch/Getty Images)

Four States Take Meta to Trial in California, the Latest Front in Sprawling Multidistrict Fight

Vietnam Tightens Online Speech Controls as Cyber Law Takes Effect

Cemetery for people on the move in Soufli, Evros, Greece. Photo by Petra Molnar.

The Racial Necrotechnics That Lead to Death By Design at Automated Borders

AI Meets the US Midterm Elections

Michigan residents turn out to vote in the midterm elections at Brookside Christian Reformed Church in Grand Rapids, Mich., on Tuesday, Aug. 4, 2026. (Isaac Ritchey/The Grand Rapids Press via AP)

How AI Is Reshaping Election Information Ahead of the 2026 US Midterms

A brightly coloured illustration which can be viewed in any direction. It has several scenes within it: people in front of computers seeming stressed, a number of faces overlaid over each other, squashed emojis and other motifs.

Here Come the Synthetic Voters

A person passes next to an Aadhaar Seva Kendra in Kolkata, India on August 15, 2023. (Photo by Sudipta Das/NurPhoto via AP)

Aadhaar's Success Story Deserves a Closer Look

Amber Sinha is a contributing editor at Tech Policy Press.

In his annual Independence Day address on August 15, Indian Prime Minister Narendra Modi 

 that over ten million young people will receive AI training through a dedicated national drive over the coming year, while India’s digital public infrastructure will be harnessed to deliver free online test preparation for competitive exam candidates. This follows the government’s announcement in late July about the 

constitution of a high-powered task force

 to recommend comprehensive steps to make India’s examination system “leak-proof and technology-driven.”

These steps come on the heels of the unprecedented Cockroach Janta Party (CJP) protests that recently swept the country, driven by a massive 

 that exposed the severe vulnerabilities of India's centralized examination infrastructure. As millions of young people took to the streets and demanded accountability for compromised futures, the political fallout culminated in the 

Faced with a structural and political crisis of this magnitude, this response is revealing. The framing of the task force 

 the issues with the examination system as 

, a problem of insecure databases and inefficient data pipelines. 

This represents a limited and extremely technocratic view that has informed education policymaking in India over the last decade. It is an approach that consistently treats pedagogical, structural, and socio-economic failures as mere data management and fraud detection problems, applying digital band-aids to a system that is fundamentally misaligned with the needs it seeks to address.

A decade of digital interventions and the employability crisis

To understand the current moment, one must look at the trajectory of education and examination reforms undertaken by the Modi government over the past decade. The state’s vision has been to leverage Digital Public Goods (DPGs) to achieve scale and efficiency. The last ten years have seen an aggressive push toward the digitization of the classroom and the examination hall.

DIKSHA (Digital Infrastructure for Knowledge Sharing)

, envisioned as a national platform for school education. This was followed by the Central Board of Secondary Education’s (CBSE) implementation of the 

, designed to digitize answer scripts and streamline evaluation to prevent manual tampering and reduce logistical bottlenecks. The government also introduced the 

APAAR (Automated Permanent Academic Account Registry) ID

, marketed under the slogan “One Nation, One Student ID.” Concurrently, there has been a significant push to integrate AI into educational delivery and assessment, promising personalized learning and automated grading.

This digital architecture often operates divorced from the material realities of the Indian education system. The technological reforms have done little to address the most glaring failure of this system: the employability crisis. India’s much-touted demographic dividend is rapidly transforming into a demographic burden. Successive iterations of the 

data from the Center for Monitoring Indian Economy (CMIE)

 have repeatedly demonstrated that nearly half of all Indian graduates remain unemployable in the modern economy, lacking fundamental critical thinking, problem-solving, and communication skills.

The structural issues are glaring — a massive shortage of qualified faculty, chronically underfunded state universities, rote-learning-based curricula, and a severe lack of physical infrastructure. In this context, the government’s flagship 

, launched to bridge the gap between education and industry needs, has largely failed to meet its ambitious targets. 

The focus has consistently remained on certification and enrollment numbers rather than pedagogical quality or actual skill acquisition. Instead of addressing these root causes — which require sustained financial investment and complex institutional reform — the state has prioritized digital legibility. By focusing on metrics, dashboards, and technological interventions, the government has created an illusion of progress, while the fundamental unemployment problem and the quality of education remain disastrously stagnant.

The current approach to education reform represents a specific, highly technocratic ideology. It is a governance model that operates on a fundamental premise of trust deficit. In this worldview, students, teachers, and regional administrators are inherently prone to inefficiency or corruption, and centralized technology is the only objective, incorruptible arbiter. Consequently, the primary problem statement that educational technology seeks to solve is not how to improve learning outcomes, but how to optimize fraud detection and standardize assessments.

This ideology is most evident in the design and deployment of the APAAR ID. The state frames APAAR as a revolutionary ‘single source of truth." The 

 is to replace fragmented, physical academic files with a unified, lifetime digital history, ostensibly to stop the proliferation of fake certificates and eliminate duplicate records via frictionless digital verification. However, this top-down architecture replicates the exact ideological and structural flaws of the Aadhaar project. By demanding absolute centralization and standardization, the system becomes inherently brittle and exclusionary.

This uncritical embrace of technology is equally evident in the state's recent push toward Artificial Intelligence in schools. As a recent investigation highlighted, Indian schools and state portals are embracing AI without adequate guardrails, precipitating a 

" Ed-tech platforms integrated into state systems and automated AI tutors are generating factually incorrect information, while automated grading systems penalize students whose subjective answers do not align with the algorithmic vector space. 

AI is being deployed not to augment teacher capacity, but to substitute it, driven by a desire for cost-cutting and scale. The risks of algorithmic bias — where AI systems trained on skewed datasets systematically 

 students from specific socio-linguistic backgrounds — are entirely ignored in the 

state’s rush to present a technology-driven educational facade

It is within this context that the appointment of Nandan Nilekani, among the chief proponents of India’s modern technocratic state, to run the task force must be seen. His trajectory—from building Aadhaar at UIDAI to his influential role at the software product think tank iSPIRT, and his championing of India Stack and Digital Public Infrastructure (DPI) — represents a very particular philosophy of governance.

This governance approach has so far reduced complex, deeply human civic and social interactions to APIs, databases, and authentication protocols. It operates on the belief that societal problems are essentially engineering problems at scale. If welfare delivery is inefficient, build a biometric database (

). If retail commerce is skewed, build an open network (

). Now, if the examination system is plagued by leaks and protests, will the response focus on building a new layer of cryptographic or biometric verification?

This model of building digital public infrastructure fundamentally alters the relationship between the citizen and the state. In this framework, citizens are reduced to data points, and systemic issues of equity, access, and democratic accountability are bypassed in favor of technological efficiency. The architecture of 

 and similar Indian state-endorsed DPI is inherently centralizing. While it is presented globally as a benign, open-source public good, it allows the state to consolidate vast amounts of behavioral, financial, and now academic data, creating monopolies of information under the guise of public utility.

governance of Digital Public Infrastructure

 has noted, this infrastructure is routinely built and deployed outside of robust legislative frameworks. The architecture precedes the law. Systems are deployed via executive mandate or public-private partnerships driven by entities like 

, leaving citizens with little democratic recourse when systems fail or data is compromised.

The experience with ID and DPI systems built and championed by Nilekani has historically prioritized state fiscal savings over basic survival rights, narrowing down to a specific and limited problem statement (identification, KYC, data exchange layers) and often proposing datafication of human interactions and behavior as a key to solving complex problems.

The constitution of the high-powered task force under Nilekani is a signal that the Indian government intends to double down on its techno-solutionist approach. What is more unlikely is a recommendation for the massive expansion of the education budget, the widespread hiring and rigorous training of teachers, or a fundamental overhaul of a pedagogical model that prioritizes rote memorization over critical inquiry.

The state will undoubtedly build a more sophisticated dashboard, and the examination system may even become more “leak-proof.” But a secure database and digitalization of student data do not lead to better education. Until the government recognizes that the crisis in Indian education is structural and economic rather than merely technological, millions of graduates will continue to emerge from this seamlessly integrated digital pipeline fundamentally unprepared for the world.

Audio of this conversation is available via your favorite podcast service.

 titled “How ideas of a vast censorship network moved from the online fringe to Trump policy.” It considers the emergence and impact of the narrative that revolves around the idea of the “censorship industrial complex,” and how a network of aligned right wing actors collaborate to build that narrative. It also details the impact on civil society groups and independent technology researchers who have been targeted with lawsuits, investigations, visa restrictions, and various forms of harassment for their work.

This week, I spoke to two of the authors of the piece about what they learned reporting it, and where the story might go next: 

, MIT Technology Review senior reporter for features and investigations, and open source investigative reporter 

What follows is a lightly edited transcript of the discussion.

Eileen Guo, MIT Technology Review, where I am a senior reporter for features and investigations.

I'm Gisela Pérez de Acha. I'm an OSINT investigative reporter.

I want to thank the two of you for joining me today to talk through 

 that you produced on the censorship-industrial complex. We're going to talk through some of the details of the piece and some questions that I have, after both reading it and following this issue for a few years.

But let me just ask to start… nine months of reporting, more than 100,000 social posts scraped, hundreds of thousands of pages of records. Give listeners just a shape of the scope of the undertaking.

What were you looking at? What made you decide that it warranted this effort in this moment?

Maybe I'll just start with how this story came to be. It really came about, actually, even before those nine months that we've been investigating this.

Last April, I got a tip that this small office at the State Department that focused on monitoring and countering foreign disinformation was going to be 

. And that that office had long been accused of working at the center of this so-called censorship-industrial complex.

And I'm not a disinformation beat reporter, so I was really coming at it from my limited perspective. Important office is being shut down in the midst of all of these other DoD shutdowns. What's going on?

And the mention of this idea of the censorship-industrial complex, and the fact that it was being announced in an interview with a person that most people haven't heard of, including me at the time, named Mike Benz, really made me want to understand what's actually behind this.

How did it come to this moment that we're at? And just again, in the midst of everything else that's been happening since January 2025, where could this be going?

I want to get a little more into the implications, the effect that state and other parts of the government, for instance, but I want to stick a little bit just with the genesis.

I mean, I think one thing that's special here, you try to go back to the ideological origins as well as the events that led to what we can now think of as this meme around the censorship-industrial complex.

Walk me through that genesis. You trace language like ‘censorship industry’ on Twitter back to more than 15 years ago, and then you bring us up to date, the 2022, 2023 timeframe. What's the through line from 2011 to 2023 to today?

I'll let Eileen walk us through the through line, but I think it's really important to say that a lot of these conspiracy theories have been enacted as policy in the Trump administration, and it was important for us to figure out where this came from.

It was a massive effort. And one of the biggest challenges that we faced was actually that after Elon Musk bought Twitter, he ended all the free access to Twitter's API and making it prohibitively expensive for many researchers, us included, to really gain a full picture of what's happening in the platform.

A few years ago, I would just go to the platform and scrape and just get all of the censorship tweets. And now we just had to go around it and find creative ways and work with what we had, given all the roadblocks that Elon Musk had put for us.

Yeah. And so we ended up looking at, I think, around 100,000 tweets, but had we had full access, we could have looked at a lot more.

And to tie that back to your question, I think we could have started this story also at so many different points because this narrative of censorship has been going on for a lot longer.

In our story, we really were focusing on three main historical events that led to the sense that maybe there was censorship in quotes going on. And I actually think that it's really important to talk about this idea of censorship and what is and is not censorship.

Because the way that we use it colloquially is just, it's not very precise and it's very different from the legal definitions of censorship, which are also not really defined in US law.

But the closest is the First Amendment protections of speech, which is specifically talking about the way that the government might be quote unquote "censoring" again, people.

So that's really important because a lot of what people that have been claiming that there has been a censorship industry, or a censorship-industrial complex, or the other types of phrases that are used are really conflating these ideas of legal censorship versus what is often happening, which is content moderation.

And so the three historical events that we start out the story with, and we discuss this, we call it in our story, the origins of this discontent, are in 2016 when Russian influence operations were pretty widely recognized as having a really big influence in both the US presidential elections as well as the UK's Brexit vote.

Obviously, there are a lot of events that led up to this, so it wasn't just this but it was this really important point and transition from this earlier period where everyone thought that the internet was going to be great for democratization and both of voices.

Anyone can have a voice as well as just the democratization process around the world to, "Well, actually it's not all that rosy and we need to think about content moderation and safety on the internet maybe in a more proactive way than we've done in the past."

And then in 2020, there was obviously the pandemic, which led to a lot of rumors and misinformation and just again, this period where there was a lot of distrust, there was a lot of polarization and different ideas about what we should be doing and what we shouldn't be doing.

And in both of these time periods, we had these big tech platforms that were taking on a much more active role in moderating what was on their platforms and also talking to government and other experts because they were recognizing that they had first let Russian disinformation or Russian information operations proliferate for far longer than they should have, and that something needed to change.

And then finally, the other thing that we talk about is in 2021 with the J6, the insurrection, which was itself driven by false claims that the 2020 election had been stolen. The sense that platforms needed to do more led to President Trump being removed out of fears that he was going to be stoking more violence.

And so all of these things led to this sense that there are certain viewpoints that are being removed at greater levels.

I want to get into what you've found about that, the extent to which those claims hold up. But I actually want to ask you just to maybe steelman the censorship-industrial complex argument a little more.

Out of that origin that you just described, where is the core concern? Where do you locate it? You talk about the grain of truth.

I think there are a lot of folks out there that may not necessarily agree with any of the details of the proponents of the censorship-industrial complex narrative, and yet they're sympathetic to the idea that there are reasons to be concerned that government, or platforms, or maybe some cabal of academics are somehow coordinating on content moderation in a way that's affecting politics.

How do you think about the most charitable, the best version of the argument that those folks are making or buying into?

That's a really great question. I mean, I think that's such a great point because there is this widespread concern that is not partisan, that big tech and social media platforms have a lot of power that is often not very well explained about the content or the accounts that they are either removing or algorithmically boosting or not.

So I think that is something that is shared and with AI is increasing in new ways. But the censorship-industrial complex, this argument I think is actually more specific.

It's the idea that academics, civil society organizations, and big tech have all been conspiring to specifically remove conservative or populous speech and really at the behest of the Biden administration.

And that they're hiding their real purposes by saying that they're fighting disinformation, but then the people that believe in this or that are promoting it are seeing that again, to them at least, it seems to be affecting speech that they agree with.

I think another way to look at this or another way to explain this is content moderation decisions that at the time might've seemed right, acceptable, necessary, what we call that grain of truth. Information about the Hunter Biden laptop, this laptop that had alleged corruption information on Joe Biden and Ukraine, that was taken down.

And Jack Dorsey from Twitter later admitted that was a mistake. Mark Zuckerberg also talks about that. And that's something that the right has really held onto as the basis for them to prove the censorship-industrial complex.

And what's really fascinating as we were reporting this is the best conspiracy theories have that grain of truth. They have that content that was taken down, whether we want to call it censorship or not, whether social media companies are actually able to censor a state or not, that content was taken down.

And that's where the spiraling of the conspiracy campaign and the propaganda campaign starts. From that one thing that gets spun out of proportion into a cabal of academics, government, and big tech conspiring against right-wing speech.

I think that's right. It's this idea that when something happens, then it gets applied to everything else. And that's the problem that we go from the actual facts to this is the vibe and policy is being made on vibes.

And then I think the term itself, it's so catchy. It's like if this were a marketing campaign, it would've been a right-on term to pick.

Because it comes from the military industrial complex, which is something that both the left and the right use a lot, feel very comfortable with in describing these alliances, I guess, between academia and government, an industry that brings something forward.

So the censorship-industrial complex has such a great ring to it already that I think that also contributed to the virality and how easily it was adopted through the internet.

Well, since you've used the term marketing to talk about this, maybe I'll ask you a little bit about where we've got to in more recent years and the extent to which pushing this idea of the censorship-industrial complex has on some level become a business.

Has become something that is supported by big money and is making money for the people who are pushing it. What did you find when it came to those types of questions?

We found that there was this really big network of media organizations, of grant-making organizations, and then individual influencers like Mike Benz, who was the one 

 Marco Rubio, which we talk about in the start of the story that kicked all of this off.

And it's millions of dollars that are supporting various organizations that are behind this. And so it has become a really big business, and I can go into the network of how that looks if that would be helpful as well.

Give us some of the basics. What does it look like? How would you sketch it out for a listener who's unfamiliar with this network?

So after President Trump's first term in office, a number of his former administration officials started these media and advocacy groups.

And so I mentioned Mike Benz, he started a group called the Foundation for Freedom Online, and we found him to be really the key player that has been talking about censorship. Since before he was using his own name, he used to have a podcast called Frame Game Radio where he was anonymous, but he was talking about censorship as primarily a Jewish phenomenon.

He was talking, at the time, about the ‘diversity-industrial complex’ and how all of these networks really needed to be taken down in order for, at the time they were talking about the alt-right still, so in order for the alt-rights ideas to be popular.

So there was Benz who served in the Trump administration for a very short period of time. There was Stephen Miller who started a group that became a really key part of all of this, America First Legal, which they pop up in a lot of other places.

But in the free speech and censorship space, they got a lot of the documents from the government that was then used by Benz and the Foundation for Freedom Online to make the argument and to provide the citations.

And then there were the media organizations, and in particular, John Solomon's Informing America Foundation that was giving money out to right-wing podcasts, including Charlie Kirk's podcast, John Solomon's podcast, Steve Bannon's podcast. They're all supported by these same networks.

And so we just found that they were mutually reinforcing and putting forth this idea that there is massive amounts of censorship against the right. And from there, these ideas kept getting picked up and cited in more mainstream places and then moving into policy. And that's where it really became dangerous.

What I think is also particularly interesting is that when we first set out to map the origins of the censorship-industrial complex, we thought that we might find a bunch of bots amplifying certain messages.

And when we actually got into the data analysis, it didn't look like bots at all. It looks like what you would've called organic content, but by what Eileen just described as a really big apparatus that is used strategically to amplify each other.

And they use it not only against. In the piece we mentioned the example of Cat Park, this video game that was used to train younger folks on spotting disinformation. And you see this playbook work against Cat Park in a really clear way.

First, the game comes out, then Stephen Miller's AFL files FOIA requests to get documents on the game. Then once they publish those documents, Mike Benz takes those documents and posts them on his organization.

And then after that, this entire Informing America Foundation ecosystem amplifies whatever Benz and AFL have put out, and then there's a congressional hearing where you mentioned Cat Park.

And it also has to do with the fact that the House became majority Republican, and that also gave a lot of political power to pick up these sorts of topics like that.

Yeah, I mean, I think that's one thing that we followed closely at Tech Policy Press over the last couple of years, Rep. Jim Jordan's (R-Ohio) committee in particular and the weaponization effort that they've led there to try to dig into this.

And it does seem like it's that kind of a cycle of we will find some new set of documents or unearth some new thing that we can hold up to the light.

And you see that cycle begin where some citations created by one of the characters in this network, it ends up in front of the committee, and then there's another round of discourse and attention to it after it gets surfaced in the committee.

I have to say that the last couple of hearings have really felt like a recycled concern. It's been difficult to, I don't know, get the sense that there's a lot of new material. I don't know. Do you have that sense at all that they are recycling the greatest hits rather than being able to churn out much that's new?

Yeah, absolutely. I mean, if you look at Mike Benz's Instagram page, which I follow still, he's still talking about USAID, which was dismantled. He's still talking about the same things that they have been talking about for years. So I think that's right.

I want to ask you just a little bit now about the wreckage, some of the implications. I mean, you've already mentioned the R/FIMI or R/FIMI. There's things like the Global Engagement Center, CISA, the effects on the FBI. You mentioned USAID. Some of these are very consequential in terms of what the effect has been.

I find it very ironic on some level that some of the cuts at USAID have cut actual anti-censorship efforts that the United States government was engaged in abroad, working in Asia, working in Latin America, elsewhere, efforts to support digital rights activists and others who are contending with real and true censorship by authoritarian regimes.

But what else would you point to as a particular, I don't know, forms of impact that this narrative and the political power that's come behind it have led to?

What was really striking is to see this playbook in action, for example, against the Disinformation Governance Board, which we talk about in the story.

The shutdown of other disinformation-focused organizations like the Stanford Internet Observatory, the Election Integrity Partnership, and later government agencies like R/FIMI as well.

In an aerial view, the 1,000 acre Meta El Paso Data Center as of August 13, 2026 in El Paso, Texas. (Photo by Brandon Bell/Getty Images)

 is this: “superintelligence-for-all” empowerment will never happen if Meta and other hyperscalers monopolize compute. Control over compute is a tool of domination as powerful as any of the frontier model secret weights that Zuckerberg slags.

 observes wryly that in the manifesto, “a man who single-handedly commands a $1.5 trillion tech giant comes out as a critic of corporate power.” Platformer’s 

 notes that while Zuckerberg “celebrates individual voices and democracy,” his company “is working to warp the democratic process through secretive data center deals.” Through those deals, Meta is trying to catch up with the other four hyperscalers (Amazon, Google, Microsoft, and Oracle), who together 

 of the world’s AI compute. Zuckerberg doesn’t talk about 

 source of power. Instead, he focuses further up the stack on models and applications. Here, he says, “[r]ather than centralizing superintelligence, we should distribute it widely and give every person the ability to direct it.”

 the physical infrastructure needed to produce that superintelligence? Infrastructure power 

, in its 2024 report on open weight models (p.30) noted that access to compute is a significant chokepoint for innovating with such models. As 

, “the computational resources needed to build new AI models and use existing ones at scale… are scarce, extremely expensive, and concentrated [and] this significant resource asymmetry undermines any claims to democratization that the availability of ‘open’ AI models might be used to support.” (p.8)

In other words, frothing about AI democracy and liberty is useless without addressing compute. Zuckerberg knows this. That’s why he acknowledges compute as a pinch point in the flow of freedoms he champions. “No matter how intelligent AI becomes, there will always be a finite amount of compute and therefore an opportunity cost for how we use it,” he writes. His response to this scarcity is for the tech giants to build ever more compute so that there’s enough for training “recursive self-improvement” while also “committing the significant majority towards people's individual goals.” Here, as elsewhere in the manifesto, governance choices are the tech companies’ to make. 

 will own the means through which people self-actualize. Every “exceptionally capable personal agent” and every “powerful tool to build businesses” and all the other compute-driven wonders the manifesto teases will depend on the few companies controlling compute.

For the conveniently libertarian political theory of Zuckerberg’s manifesto, this concentrated private power is unimportant. The only real threat to AI-enabled individual liberty is government tyranny. “There is a risk that an imbalance of power between individuals and government leads to a loss of freedom or totalitarian state,” Zuckerberg writes. “This is a constant tension in democracy.” What’s also a constant tension in democracy is how to constrain overweening private power. Technology policy in the American democracy has never been libertarian. Zuckerberg’s only recognition of private compute power is his assurance that Meta (and other benign titans) will distribute compute widely and focus heavily on alignment. In other words, “trust us.” Given Meta’s history on the trust front, as it appeals a pathbreaking New Mexico court judgment that it is a public nuisance, this tack is laughable. And it has been loudly laughed at 

. Obviously, binding and transparent regulation is a more effective check. The manifesto will have none of it ,and Meta has been instrumental in making sure the federal government enacts none.

What might a compute policy look like that took self-empowerment and democracy seriously? A person writing a version of Zuckerberg’s screed with an eye to network studies might have cited Ithiel de Sola Pool’s classic 1984 work, 

. Pool wrote about how new technologies (like “electronic publishing”) might usher in new freedoms through “distributed intelligence” (p. 251) if only government could abstain from centralizing control. This fits well with Zuckerberg’s paean to decentralized open weight models and preference for policy that amounts to little more than facilitating Meta’s business model. But here’s the thing. Pool understood that freedom at the application layer might require regulation at the infrastructure layer in the form of common carriage and access requirements.

For compute, and more generally for cloud services that mediate access to compute, we can look to the same family of tools long used in US regulation of concentrated networks, platforms, and utilities.

Remove chokeholds on access to compute where necessary through some combination of nondiscrimination, interconnection, and common carrier obligations. Jai Vipra and Sarah Myers West of the 

 have a nice report on this from 2023 when the FTC and the Department of Treasury were beginning to sound the alarm on undue concentration and points of failure in the compute and cloud services markets. Last year, 

 of the Vanderbilt Policy Accelerator published a blueprint for federal legislation for the AI stack, including compute, that he described as pro-competition and pro-national security, and which might actually create the balance of power Zuckerberg’s manifesto praises.

Ensure that there is some meaningful supply of “public compute” that is not controlled by any company. The scarcity of compute for scientific research has led to nascent federal efforts to amalgamate commercial and government compute for research in the 

 with new compute resources, again mostly for research. One might imagine a federated network of state compute forming a sovereign compute infrastructure of the kind 

 is trying to build. A new idea would be to regularize public compute set-asides as part of data center builds. What if hyperscalers had to set aside a certain amount of  “

”, like a public park, for the autonomous use of nonprofits, local governments, and educational institutions? Cable companies had to set aside PEG channels when building out their networks and the FCC reserved noncommercial broadcast channels for public service.

Require transparency. One could talk about transparency at each of the layers in the stack, but let’s limit it here to compute. Here too, the libertarian Pool acknowledged that to have freedom at the top layers – the only ones Zuckerberg really addresses – we need to know how companies are managing the infrastructure choke points. In the case of compute, this might include the amount of compute capacity a firm actually controls, to whom that capacity is allocated (including to the owner itself), and pricing terms.

For ten years at least, Zuckerberg has repeatedly promised connection and democracy on his platforms. Instead, we’ve gotten slop and oligarchic control. The AI manifesto promising superintelligence-for-all is cut from the same cloth. But still, it’s worth taking the aspiration at face value and asking what it would take to actually democratize AI at each layer of the stack, starting with compute. It would require that people who want to run models and applications can do so freely without control or interference from infrastructure providers.

Sarah Saadoun is a senior economic inequality adviser at Human Rights Watch.

 about the economic fallout of artificial intelligence have brought tax policy into the spotlight. Governments could 

find themselves desperate for more revenue

 at exactly the same time that they are losing the taxes those vanished jobs would have generated.

There is no magic bullet governments can use to resolve this problem. One approach with 

 AI profits. The idea is that governments could use that revenue to help manage the human rights impacts of rapid, AI-driven change. This approach has support in some surprising quarters. 

 have both floated proposals that would increase taxes on companies benefiting from AI as part of their case for widespread adoption of AI.

The current international tax system doesn’t allow states much room to do this effectively, however. A century-old rule common in bilateral tax treaties requiring physical presence 

 of the world’s governments, which was a growing problem even before the advent of AI. A change requires a much larger, collective solution to the larger failings of our global tax system. Luckily, it is already taking shape.

 that could finally make it easier to tax tech companies, among other badly needed global tax reforms. Not only could this treaty better equip governments to weather AI-driven uncertainty, but it could give them essential tools to better protect and realize human rights across the board.

Rules for cross-border taxation set out in the 1920s and fine-tuned by the Organization of Economic Cooperation and Development (OECD) form the basis of thousands of bilateral tax agreements between countries. These treaties generally let governments tax a foreign company’s profits only if it has a “permanent establishment”—a physical presence—in the country.

This means tech giants like Meta or Uber, which often operate entirely remotely, can earn fortunes in a country while paying tax on profits somewhere else entirely. For instance, Facebook has some 

Latest

Podcast

Recent

Our content delivered to your inbox

Thank you!

Around the globe