If you’ve had the sneaking suspicion that the internet is worse than it’s ever been, you’re not imagining it. Over time, your favorite app gets cluttered with ads or a platform you’ve invested years into suddenly locks you out or changes the rules. That’s all part of a predictable pattern. Cory Doctorow calls it “enshittification,” and his book, Enshittification: Why Everything Suddenly Got Worse and What to Do About It, is the most important tech policy book I’ve ever read.
As someone who’s spent my career in tech, I now watch the AI revolution with alternating awe and horror as it accelerates everything we do in every arena. My heart breaks as democracies around the world fall and biases we long thought were dead surface again. At this pivotal moment, I believe this book should be required reading for everyone in tech and every politician making decisions about technology regulation. We can fix what’s been enshittified and Doctorow has a plan.
The Four Phases of Enshittification
What is “enshittification”? You’ll intrinsically recognize it in your favorite online platforms once you hear Doctorow’s simple framework. Enshittification follows a predictable lifecycle that plays out again and again across digital platforms:
Phase 1: Be Good to Users. The platform starts by providing an excellent service. Think early Facebook, when it was just your college friends sharing photos. Or Amazon, when it was a revolutionary way to get any book imaginable delivered to your doorstep super fast. At this stage, the platform loses money or breaks even, but it’s building something genuinely useful.
Phase 2: Abuse Users to Be Good to Business Customers. Once the platform has locked in a critical mass of users, it starts to squeeze them, preferring money-generation over customer value. Facebook’s algorithm stops showing you posts from your friends and instead pushes sponsored content. Amazon’s search results get clogged with sponsored products, and “honest” reviews are gamed by sellers. The platform is now optimizing for business customers (advertisers, sellers) at the expense of users.
Phase 3: Abuse Business Customers to Claw Back Value For Yourself. After the business customers are locked in, the platform turns on them too. Facebook charges more for ads that reach fewer people. Amazon increases seller fees and steers customers toward its own house brands. Now the platform is extracting maximum value for its shareholders.
Phase 4: “A Giant Pile of Sh*t”. Eventually, the platform becomes so terrible that users and business customers flee to alternatives. Except when they can’t, because the platform has used anti-competitive practices to eliminate alternatives and lock everyone in.
This isn’t a conspiracy theory. It’s a business model, and it’s everywhere.
Who Has Enshittified?
Doctorow’s book has four detailed case studies for companies and products that have run through all four phases already: Facebook, Amazon, iPhone, and Twitter. But, you can recognize this pattern in a lot of your beloved tools. Anyone who is forcing you to accept new terms so they can mine your data to help train their AI Models, are leaning in this direction and most companies are getting away with it. Some, like Adobe, have gotten so much pushback that they’ve lost revenue and switched tactics. Most, however, are happy to streamroll past their customers as fast as they can with their revenue-hogging train.
I learned a lot of things in this book that I never knew. The one that shocked me the most is that Amazon actually causes prices to increase across the internet. Amazon’s agreements with Prime vendors require them not to sell their products more cheaply outside of Amazon. Amazon, meanwhile, takes 45-51% of every sale, causing sellers to raise their prices on Amazon to cover the cost and because they can’t sell their products more cheaply elsewhere, they raise prices everywhere. You know as well as I do that you can find damn near everything on Amazon today. So, imagine the influence they’re having on the prices of everything we buy.
No wonder it feels like we’re being squeezed on all fronts right now.
Why This Matters Now: The AI Wave Is Coming
If you think enshittification is bad with social media and e-commerce, wait until it happens with AI. We’re already seeing the groundwork being laid.
The major AI companies are following the same playbook. They’re starting in Phase 1. Right now, they provide amazing tools, often for free or cheap. Claude, ChatGPT, and others offer capabilities that seem almost magical. Users are rushing to adopt them, building workflows and businesses that depend on these tools.
But what happens when these companies move to Phase 2 and 3? When they raise prices dramatically? Or they use the data from your queries to train models they sell to your competitors? When they decide your use case isn’t profitable enough and shut off your access? When they leverage their market position to prevent competitors from emerging?
AI is different from social media in a crucial way: it’s infrastructure. If Facebook enshittifies and you leave, you lose contact with distant relatives. If your AI infrastructure enshittifies, your business might collapse. The lock-in is deeper, the switching costs are higher, and the potential for abuse is greater.
This is why we need regulation now, not after the enshittification cycle is complete.
The Four Things We Can Do to Fight Enshittification
Here’s the good news: Doctorow doesn’t just diagnose the problem, he provides a roadmap for fighting back. And every single one of these solutions requires us to be more open to regulation, which is something the tech industry has historically (and successfully) resisted.
1. Break Up Monopolies and Prevent Anti-Competitive Mergers
Big Tech’s power comes from market dominance. When a platform controls a market, it can abuse users and business customers because they have nowhere else to go. The solution is straightforward: enforce antitrust law aggressively.
This means blocking acquisitions designed to eliminate competitors (like Facebook buying Instagram and WhatsApp). It means breaking up companies that have become too dominant. It means preventing platforms from using their market position in one area to advantage themselves in another (like Amazon using seller data to launch competing products).
For AI specifically, this means careful scrutiny of partnerships between AI companies and cloud providers, of training data monopolies, and of companies acquiring AI startups to eliminate competition. It also means taking a closer look at our existing tech giants and not being afraid to break them up.
2. Mandate Interoperability
One reason you can’t leave Facebook is that all your friends are there. One reason you can’t leave Amazon is that’s where all the sellers are. Platforms deliberately make it hard to take your data and relationships elsewhere.
The solution is to require interoperability. If you could use any social media client to talk to people on any social network, Facebook couldn’t hold you hostage. If any e-commerce platform could list Amazon sellers’ inventory, Amazon couldn’t abuse them as easily.
For AI, this means requiring open APIs, standardized data formats, and the ability to move your data and custom models between providers. It means preventing vendor lock-in through technical and legal requirements.
3. Protect End-to-End Encryption and Privacy
Enshittification depends on surveillance. Platforms need to know everything about you to manipulate you effectively, to know how far they can squeeze before you leave, and to extract maximum value from your data.
Strong privacy protections and end-to-end encryption make enshittification harder. If platforms can’t spy on users comprehensively, they can’t abuse them as precisely. This is why Big Tech fights privacy regulations so hard—surveillance is central to their business model.
For AI specifically, this means regulating what training data can be collected and how, requiring transparency about data usage, and giving users control over their information. It means preventing AI companies from building comprehensive surveillance profiles under the guise of “improving the model”.
4. Protect Workers’ Rights to Organize and Speak Out
Some of the most important whistleblowers about tech platform abuses have been employees. But companies use non-disclosure agreements, non-compete clauses, and the threat of losing stock options to silence workers.
Protecting workers’ ability to organize and speak out about corporate misconduct is crucial for accountability. This includes reforming NDAs, limiting or eliminating non-compete agreements, and protecting whistleblowers.
For AI companies, this is especially important given the potential safety risks. Google employees speaking out have helped reverse several of the company’s bad decisions. Tech workers are good people who, in general, care about the right things. We need AI researchers and engineers to be able to speak publicly about concerning capabilities or business practices without fear of retaliation.
Why Technologists Need to Support Regulation
Here’s where I need to challenge my fellow technologists. I know the instinct is to resist regulation. We’ve been conditioned to believe that regulation stifles innovation and that the tech industry should self-regulate.
But look around. Self-regulation has failed. The platforms have enshittified. The internet is worse than it was ten years ago. And without intervention, AI will follow the same path. We’ve all seen our executives make decisions that we don’t agree with and feel the sting of not having enough power to reverse those decisions.
The regulations Doctorow advocates for—antitrust enforcement, interoperability requirements, privacy protections, and worker protections—don’t stifle innovation. They prevent the kind of market consolidation that actually stifles innovation by eliminating competitors.
Would the internet be better if there were a dozen social networks competing on features and user experience instead of Facebook’s monopoly? Would e-commerce be better with more competition? I 100% think AI is better because we can switch between providers based on quality rather than being locked into whatever vendor they started with. Just look at what happened when I tried to form teams for my Capstone course with Claude vs Gemini.
Regulation done right doesn’t hurt innovation. It prevents the people who innovate first from pulling up the ladder behind them.
Why Politicians Need to Read This Book
If you’re a policymaker reading this, Doctorow’s book provides the framework you need to understand what’s happening in tech and what to do about it. The tech industry will tell you that regulation is too hard, that technology moves too fast, that you can’t possibly understand it well enough to regulate it.
That’s self-serving nonsense. We regulate airlines, pharmaceuticals, and financial services and all of them are complex, fast-moving industries where mistakes can be catastrophic. And, we can regulate tech. We must regulate tech.
The specific policy recommendations in Enshittification are clear and actionable. You don’t need to understand machine learning to know that companies shouldn’t be allowed to acquire all their competitors. You don’t need a computer science degree to mandate that users should be able to take their data with them when they switch services.
What you need is the political will to stand up to Big Tech lobbying and to act before the enshittification cycle is complete. Because once a platform has fully enshittified and eliminated all alternatives, it’s much harder to fix.
Should You Read This Book?
Yes, absolutely. Whether you’re an engineer, a policymaker, or just someone who uses the internet and wants to understand why it keeps getting worse, Enshittification explains what’s happening and what we can do about it.
The book is accessible, you don’t need a technical background to understand it. My husband, a budget analyst turned life coach, is reading it now and getting just as much out of it as I did. It’s well-researched and rigorously argued. And it’s urgent. The choices we make in the next few years about AI regulation will shape the internet for decades to come.
And yes, I did write this with the help of Claude, but I stand by the fact that AI needs to be regulated. I think we can have our cake (the amazing capabilities of AI) and eat it too (AI not being enshittified).
Doctorow has given us the playbook for fighting back against platform monopolies. Now we need the political will to use it. And that starts with understanding the problem, which is why this book is so important.
Read it. Share it. And then let’s do something about it before the next wave of enshittification is irreversible.