Shareable does not consent to Meta’s theft of content

If you have received any emails about terms of service from a US tech firm in 2026, you may already know that the word “update” no longer means what reasonable people believe it to mean. 

“Updated,” in United States corporate parlance, means enshittified. 

Updates to terms of service (ToS) in recent years involve a number of anti-user conditions, including having to accept that a company which provides a service you rely on in your life—or are required to use for work—has decided that you will no longer have legal rights to file lawsuits or take part in class action litigation if it does anything harmful to you. 

This degradation of users’ rights typically takes the form of changes to terms around binding arbitration (for-profit dispute resolution), which the user bringing the grievance must often pay for. Binding arbitration is worse, and less just, than a court case, and that’s how these companies like it. It’s cheaper for them, in terms of legal risk and associated insurance. Just this calendar year, countless US tech firms “updated” their ToS in relation to binding arbitration clauses, including Venmo, Spotify, and Lyft, to name a few.

Companies quietly deciding that we no longer have legal recourse is the sort of degradation of civil society that we may have learned to accept, but this pales in comparison to what Meta is now trying to do.

On October 15, Nieman Lab drew attention to the fact that Meta’s upcoming “updated” ToS would likely allow the corporate giant to train its AI products on all website content from publishers with Meta business accounts. Meta is the parent company of Facebook, Instagram, WhatsApp, and Threads, among other platforms and apps. With 3.9 billion global users, this would impact an enormous number of artists, publishers, content creators, designers, photographers, videographers, etc.  Nieman Lab consulted three legal experts, who concluded that, “As written, the new terms could be used to argue that Meta can scrape content from news sites and use that content to train the company’s AI products.” 

In order to opt out, an organization with a Meta business account would not only have to delete all of its Meta accounts before the “updated” ToS go into effect and remove all Meta products from its website(s), but it is also likely that publishers would have to first go back and delete content individually from Meta-owned platforms. This is a completely unreasonable way to seek consent.

“Meta is pushing this most recent ToS enshittification so that it is on better legal ground as it steals content from artists, journalists, performers, designers, and so many more to train its corporate AI.”

This change in ToS was initially slated to go into effect on October 30. Following widespread backlash, Meta changed its proposed ToS language from “train” to “develop,” though its legal implication seems to be the same. Meta then delayed these changes to the corporation’s ToS until November 20, mischaracterizing widespread criticism of the proposed change in policy as a request by the public for further explanation, while changing little in the proposed language. Meta has since backpedaled and delayed, but offered no substantive assurances that whatever it plans on doing is meaningfully different from its initial plan.

We at Shareable understand why Meta is doing this. The New York Times has a multi-year case in court against OpenAI and Microsoft, alleging that these tech firms violated copyright by training their for-profit AI products on New York Times content. Last month, USA Today filed a similar lawsuit against OpenAI. A coalition of Canadian publishers has filed a lawsuit against Anthropic (the parent company of Claude) regarding the same sort of alleged harm: the same alleged theft of content.

Meta is pushing this most recent ToS enshittification so that it is on better legal ground as it steals content from artists, journalists, performers, designers, and so many more to train its corporate AI.

Meta knows exactly what it is doing, just as the Microsoft executive—in internal emails related to scraping content for AI training—knew what they were doing: engaging in “the largest theft of labor in human history.”

If Shareable existed in an information ecosystem that was not increasingly consolidated by billionaire owners, with little antitrust enforcement to speak of in the past few decades, this may not be a problem, and we could just leave Meta’s platforms.

However, Meta has done its best to stifle competition. In fact, Meta has bought out many dozens of startups, which it then often shuts down. Mark Zuckerberg himself stated that due to Facebook’s inferior camera app product, it chose to buy out its competition, Instagram, saying, “It is better to buy than compete.” For users of Instagram, it is clear that the platform has grown worse over the years: buggier, more harmful to the mental health of adults and children alike, and more filled with ads. 

But at least pioneering Meta can come up with ideas of its own: it spent $80 billion on the Metaverse no one asked for—more than the GDP of Uganda—only to completely scrap the failed project this year. Meta has now pivoted to peddling “pervert glasses”— so unpopular that even ICE banned their use. And now, it wants to get more into AI, which brings us back to why this ToS “update” is happening. 

Meta is a company that seems determined to buy or steal its way to victory. And when we look at the numbers, it is clear that the consolidation of attention on anticompetitive corporate platforms—not just those owned by Meta, but also Google and other tech giants—is eroding our informational ecosystem. If you have been paying attention to the financials of publishers, you will understand that Shareable is not the only one that has been deeply financially harmed by the rabid pursuit of profits by tech giants. A 2019 study found that in a single year, Google alone gained $4.7 billion from content from publishers, for which the publishers were not compensated. 

According to an analysis by Press Gazette, news consumption across news sites is down globally, and an ever-increasing share of folks are getting news directly on social media sites. As a result of the systematic reduction in traffic, starting a number of years ago, Shareable has been publishing fewer and fewer articles. Now, rather than focusing on reported stories as we used to, we have been publishing more content that enables people to do specific things in their communities and offering more hands-on services to incubate community organizers to start those projects. As visits to websites, news apps, and TV views decline, social media platforms grow and grow, consolidating more and more of the attention market. 

As such, Shareable and anyone who needs to access a broad, public audience cannot simply leave the platforms everyone stays on, no matter how harmful they are. We need the folks wanting to do good work in their communities to find us, somehow, and the tech companies that control our information platforms are increasingly hostile to that happening at all. 

“The Shareable Team looks out into the world for that which is good, and we try to amplify the good where it already thrives, cultivate it where it may grow, and provide the conditions for that pro-social behavior to replicate.”

AI summaries take content from certain places on the internet and repackage that content on the platforms of tech giants. These summaries encourage users not to click into the websites where the primary information is hosted, and instead to stay on the platform—Facebook, Instagram, Google, etc. This deprives the originator of the valuable information of revenue—revenue which would be used to create more thoughtful, deep, or original content not already on the internet. This is one form of theft: the deprivation of revenue.  

And now, the tech firms engaged in AI development wish to steal the rest of the content from those who publish useful information on the internet to train their proprietary AI models. This is theft, doubled. First revenue, then the content itself, used to develop a corporate product. 

The vision of Meta, like Google, Anthropic, and OpenAI, is one where no one who puts useful information on the internet gets paid for doing so, but those who copy and repackage that information are paid quite well. 

The difference between getting mugged versus having your valuables stolen from you by Meta is a matter of theater. The mugger is open about his threat of violence. Meta performs propriety, while letting the economic violence it has intentionally cultivated speak for itself. But what Meta steals is far greater in value than anything a mugger is likely to get: for Shareable, it is all the content we have ever created, and all we will ever create. 

With a ToS “update,” suddenly, everything we have done will all become Meta’s.

To be clear, Shareable dearly values open access to information, and we value true sharing. We want to share stories and practical knowledge about how to do things that benefit one’s community and the world. Our mission is focused on this. In fact, Shareable publishes the vast majority of our content under a Creative Commons license. This means that we allow all of our content to be republished or reprinted for non-commercial purposes, as we have for many years. Yet Meta scraping our content for their proprietary AI is a violation to all three of the parameters of our Creative Commons license—Attribution, NonCommercial, and No Derivatives.

It is only sharing if the sharer consents. Otherwise, it is theft. And we do not consent to our content being used for the development of for-profit, socially detrimental corporate AI products used in walled gardens of information. We will not leave the increasingly few platforms that we must use, due to unregulated corporate consolidation, but we do not consent to what Meta aims to do. 

The Shareable Team looks out into the world for that which is good, and we try to amplify the good where it already thrives, cultivate it where it may grow, and provide the conditions for that pro-social behavior to replicate. The absence of Shareable editorial content on poor and unethical behavior does not mean that we do not perceive the poor and unethical behavior around us. But in this instance, Meta’s actions are unconscionable, and we find it untenable to stay silent on this. 

We encourage anyone who has put anything thoughtful, informative, beautiful, or good on the internet to make visible statements about this, so that the record may show that although the law may perceive consent where there is only coercion, reasonable people can see clearly that we do not consent to being robbed by Meta.

—Aaron Fernando with Alison Huff, Bobby Jones, Paige Kelly, Tom Llewellyn, and Xaivier Ringer (the Shareable Team)

If you would like to support Shareable’s work, please consider donating. Thank you!