Discussion
After half a century of neoliberalism, we are all drenched in capitalism's established religion, the worship of property rights.
--
If you'd like an essay-formatted version of this thread to read or share, here's a link to it on pluralistic.net, my surveillance-free, ad-free, tracker-free blog:
https://pluralistic.net/2026/09/02/scrape-scrope-scrap/#collateral-damage
1/
We are so marinated in property worship that even capitalism's critics frame their critiques in "property talk," to the exclusion of other, more important rights, like human rights, labor rights and privacy rights.
To do this is to surrender before the battle even starts. Critics lose when they allow oligarchs and their apologists to choose a battlefield where they have a nearly unbeatable advantage.
2/
Take privacy: privacy is a human right, not a property right. Human rights aren't for sale. You can't sell yourself into slavery, you can't sell your kidneys to make the rent. If privacy is a property right - one that can be traded away - then Facebook's industrial-scale privacy invasions are actually fine, since you "traded" your privacy to Mark Zuckerberg in exchange for the privilege of talking to your friends.
3/
Some self-styled critics of tech monopolists say that the answer to Facebook's privacy invasions is to force the company to pay for your privacy with *cash*, rather than services:
https://www.wired.com/story/opinion-andrew-yangs-plan-to-pay-you-for-your-data-doesnt-add-up/
This is ideological capture in its purest form: the "data dividend" that Facebook would owe you under this system amounts to a few dollars per year.
4/
Some self-styled critics of tech monopolists say that the answer to Facebook's privacy invasions is to force the company to pay for your privacy with *cash*, rather than services:
https://www.wired.com/story/opinion-andrew-yangs-plan-to-pay-you-for-your-data-doesnt-add-up/
This is ideological capture in its purest form: the "data dividend" that Facebook would owe you under this system amounts to a few dollars per year.
4/
For wealthy people, the sums would be trivial, while working people, who've been on the downward leg of every K-shaped recovery for a quarter century, who've maxed out their credit cards and re-mortgaged their homes and drive Uber on the weekends to make rent, would have to subject themselves to ongoing surveillance.
5/
That surveillance is already used to determine the highest price those working people will pay - companies like Plexure inform fast food places when you've just gotten paid so they can tack an extra dollar onto your breakfast burrito in the app:
https://pluralistic.net/2026/04/30/something-must-be-done/#there-ive-done-something
6/
Being forced to sell your privacy doesn't just raise the prices you pay, it also lowers the wages you earn. The same people who can't afford this "pay or privacy" system have their private data used to calculate the lowest wage they'll accept for each ride on those weekend Uber shifts:
https://pluralistic.net/2024/12/18/loose-flapping-ends/#luigi-has-a-point
7/
In other words: not being able to afford privacy will result in you having even *less* disposable income, which will mean that you'll have to sell even *more* of your privacy. Lather, rinse, repeat.
But even the wealthy people who can afford to forego the pittances Facebook and others offer in exchange for their private information will find privacy elusive. That's because private information isn't a "rival good" - a thing only one person can own at a time.
8/
The fact that your mother is your mother "belongs" to both you and her, as well as your grandparents, your father, your siblings and your kids. The fact that *you* don't sell your family tree to a tech company won't stop all those other people from selling it on - as anyone whose foolish relations handed their genome over to 23andme can attest:
https://www.npr.org/2025/03/24/nx-s1-5338622/23andme-bankruptcy-genetic-data-privacy
9/
In the property religion, the way you can tell if something is valuable is if it has a high price. Property cultists insist that the problem with privacy is that our privacy is being sold too cheaply. They're wrong: private information isn't "mispriced" - it shouldn't *be* priced.
Human beings are the most valuable things in our world and they are literally priceless. Murder isn't "theft of life." Rape isn't "theft of sex."
10/
While insurers and civil courts have ways of calculating the "price" of an injury or violation, great care has been taken over the centuries to ensure that this does not turn human beings into commodities. You can't buy a "murder offset" that lets you kill people provided you pay into a fund that saves a human somewhere else:
https://pluralistic.net/2021/04/14/for-sale-green-indulgences/#killer-analogy
11/
Human beings are *too valuable* to be priced. We have an entire, *sui generis* way of balancing the conflicting interests of human rights. My daughter and wife have rights over me, I have rights over them, and when those rights come into conflict - say, if my daughter believes I can no longer care for myself and wants to put me in a care home - the process for resolving that conflict isn't an auction:
https://www.theguardian.com/technology/2008/feb/21/intellectual.property
12/
Your kids aren't your property. In fact, *all* the most important relationships in your life are non-market. Doctors have patients, not customers. Any time a doctor calls you a "customer" they are *demoting you*. A doctor doesn't sell you health. You have rights as a patient that far exceed the rights accruing to a mere customer. Same goes for other professions: Teachers have pupils, librarians have patrons, lawyers have clients. "Customer" is a demotion from all of these.
13/
As every "user agreement" you've ever clicked on shows, Big Tech loves to have everything defined in property terms - and so does *all* big business.
Take the fight over scraping for AI. You might think that this is a fight over the economic rights of creative workers - certainly, my fellow creative workers treat it as such. But because this debate is being framed in terms of *property rights*, rather than *labor rights*, this is a fight that workers are set up to lose.
14/
The tell here is how the media companies - who have been eroding the wages of creative workers for decades as they consolidated into a curdled, inbred oligopoly - describe the AI companies' scraping: as an unlicensed taking. Mitch Glazier, the $1.4m/year CEO of the Recording Industry Association of America issues press releases decrying AI training for image generators without negotiating a license fee first:
https://pluralistic.net/2026/03/03/its-a-trap-2/#inheres-at-the-moment-of-fixation
15/
Who's Mitch Glazier? Oh, just a former Congressional staffer who was drummed out the Capitol Building after he snuck a clause into must-pass legislation that would have transferred hundreds of millions of dollars from musicians to record labels, who was then immediately hired as the CEO of the record industry's largest lobbying group:
https://www.eff.org/deeplinks/2013/12/tpps-attack-artists-termination-rights
16/
Mitch Glazier - and the businesses he represents - aren't opposed to AI replacing artists. They're opposed to AI replacing *media companies*. Remember the Hollywood writers' strike? The proposal to replace screenwriters with chatbots didn't come from OpenAI, it came from Disney, Warner, Universal and other companies who claim that AI training is "theft."
17/
If AI training is "theft," then it can be cured by making a purchase, something that the AI companies can easily afford, thanks to the hundreds of billions of dollars they have been given by the world's richest investors, who are the high priests and cardinals of the property religion.
18/
The Hollywood writers are the *only workers in the world* who have successfully beaten back the use of AI in their workplace, and they didn't do it by making recourse to property rights. The Writers Guild is a *union* and it enjoys a weak form of "sectoral bargaining" (where all the workers in a field bargain with all the businesses at once) called "multi-employer bargaining":
https://pluralistic.net/2023/10/01/how-the-writers-guild-sunk-ais-ship/
19/
The writers' strike was an unqualified victory for the writers, who defended their *labor* rights to co-determination when it came to the use of new tools on their jobsite. Under the terms of their hard-fought contract, screenwriters don't have to use AI, but they can if they want. For example, writers on a long-running sitcom might train an AI with every script in the series' history, so they can ask a chatbot continuity questions as they beat out a new season of the show.
20/
But they don't have to do this if they don't want to, and even if they do, neither their wages nor their headcount can be reduced.
The media companies insist that scraping is a copyright violation, that it's "theft."
21/
As a matter of law, this is far from obvious or settled: the process of making transient copies of many works, performing mathematical analysis on them, and then publishing that analysis as software is *not* obviously a copyright violation, and anyone who claims otherwise doesn't understand copyright:
https://pluralistic.net/2023/02/09/ai-monkeys-paw/#bullied-schoolkids
22/
Worse: by demoting a labor rights issue to a mere property rights issue, AI critics are setting workers up to fail. Say the issue with AI training really is mere copyright. If that's so, the media companies who want nothing better than to pauperize creative workers can amend their standard contracts so that any worker who does business with them must irrevocably transfer their "AI training rights" to the company.
23/
Then, that company will *absolutely, 100%* license those rights to an AI company to create a model *designed to replace that worker*. The company will get paid for the training, and the resulting model will come with "guardrails" to stop other media companies from using proprietary data to compete with it.
24/
This is the story of the past 50 years of copyright expansion: every new copyright we've created "to help artists" was scooped up by their bosses, who grew more powerful and were able to demand more concessions from those artists, who were therefore poorer and thus needed more copyrights to help them (lather, rinse, repeat):
https://pluralistic.net/2026/08/18/enron-corpus/#SignHere
25/
If creative workers' AI fight is merely a copyright fight, then that fight can only determine whether media companies or tech companies will get the biggest portion when those workers are devoured by corporations. Only a *labor rights* fight can take creative workers off the menu altogether.
26/
Treating AI training as "theft" creates harms whose blast radius extends well beyond creative workers' livelihoods. Scraping is a *hugely beneficial* activity. If scraping - taking a vast corpus of copyrighted works without permission - is theft, then every search engine is a crime, unless it can afford to license "search indexing rights" from every site on the internet.
27/
Install bonfire.cafe
Get the full app experience
