Fibrosis is the body’s repair system turning against it. An organ takes damage faster than it can heal, and instead of regenerating working cells, the body lays down scar — collagen where functional tissue used to be. The organ keeps its shape. It looks intact on a scan. But scar tissue does not do the job of the tissue it replaced, and as it accumulates, the organ quietly stiffens until one day it fails. Hepatologists stage the progression F0 through F4. Stage 4 has a name most people recognize: cirrhosis.
I have started calling what social media has done to us social fibrosis, because no existing term captures the specific pathology. The social fabric — the web of trust, conversation, and shared reality that humans have been weaving for thousands of years — has spent two decades taking damage faster than it can repair itself. And the thing growing in its place is not connection. It is something that resembles connection the way a scar resembles skin.
The scar looks like the tissue it replaced
The resemblance is not an accident. It is the product.
Sean Parker, Facebook’s founding president, described the design brief at an Axios event in November 2017: “How do we consume as much of your time and conscious attention as possible?” The answer was a “social validation feedback loop” — a like, a comment, a little dopamine hit — that he said works by “exploiting a vulnerability in human psychology.” Then the sentence that should have ended the conversation in 2017: “The inventors, creators — it’s me, it’s Mark, it’s Kevin Systrom on Instagram, it’s all of these people — understood this consciously. And we did it anyway.”
They did it because of the business model. A tool built to connect people was funded by advertising, and advertising converts attention and behavioural data into revenue. Once that conversion is the engine, the user stops being the customer. The user is the herd — kept grazing, kept producing, milked for metadata like digital cattle. Meta’s 2024 results put numbers on the operation: $164.5 billion in revenue, $160.6 billion of it from advertising, drawn from 3.35 billion people who check in daily. In the US and Canada, the yield runs to roughly $230 per person, per year.
A platform with that engine does not optimize for whether your relationships are healthier when you log off. It optimizes for whether you log off at all. Outrage retains better than nuance. Comparison retains better than contentment. The feed learns to serve whatever keeps the tissue inflamed, because inflammation is engagement, and engagement is the only vital sign the business measures.
“The short-term, dopamine-driven feedback loops that we have created are destroying how society works. No civil discourse, no cooperation, misinformation, mistruth… We have created tools that are ripping apart the social fabric of how society works.”
— Chamath Palihapitiya, former VP of user growth at Facebook, Stanford Graduate School of Business, November 2017.
That is two of the people who built the machine, describing the machine, eight and a half years ago. Neither was speculating. Both were describing decisions they had been in the room for.
The biopsy results were on the table
The companies did not merely suspect the damage. They measured it, repeatedly, and the record of what they did next is the most damning part of the story.
In 2021, the Wall Street Journal’s Facebook Files surfaced internal research from 2019, including a slide that read: “We make body image issues worse in one in three teen girls.” Not an activist’s claim. Not a regulator’s estimate. The company’s own researchers, on the company’s own message board, two years before the public saw it.
It went further than surveys. According to a court filing unsealed in November 2025, Meta ran a study in 2020 with the survey firm Nielsen — internally dubbed Project Mercury — in which people deactivated Facebook for a week. They reported lower depression, anxiety, loneliness, and social comparison. A staff researcher wrote that “the Nielsen study does show causal impact on social comparison.” Another employee asked the obvious question: if the results stay buried and later leak, “is it going to look like tobacco companies doing research and knowing cigs were bad and then keeping that info to themselves?” Meta ended the research, internally attributing the negative findings to the “existing media narrative,” and never published them. Meta disputes the characterization — spokesman Andy Stone called the study’s methodology flawed and pointed to a decade of teen-safety changes — but juries have started ruling on the record itself. On 25 March 2026, a Los Angeles jury found Meta and YouTube negligent in the first bellwether trial over youth social-media addiction and awarded $6 million in damages; both companies are appealing. The first federal trial — six school districts against the platforms, before Judge Yvonne Gonzalez Rogers in Oakland — begins 15 June 2026, with roughly 2,000 related suits queued behind it.
The harm is not confined to individual psychology, either. Amnesty International’s 2022 report on Myanmar concluded that Meta’s engagement-driven algorithms “proactively amplified” anti-Rohingya content and that the company “substantially contributed” to the atrocities committed against the Rohingya in 2017 — after years of explicit warnings from human-rights groups on the ground. At the far end of the engagement gradient is not a sad teenager. It is ethnic cleansing with an amplification system.
This is the pattern that makes “social fibrosis” the right frame rather than just a vivid one. Fibrosis progresses when the insult is sustained — when the thing causing the damage keeps being applied. For two decades, every time the companies’ own instruments showed scarring, the response was to protect the insult.
The honest caveat, and why it cuts the other way
There is a real scientific counterargument, and it deserves to be stated properly. When psychologist Jonathan Haidt published The Anxious Generation, researcher Candice Odgers argued in Nature that the published evidence does not support claims of an epidemic caused by phones: the measured associations are small and inconsistent, causation could run in reverse, and we have moralized new media before — comics, television, video games — and been wrong. Serious researchers hold this position. It is not industry spin.
But Project Mercury reframes that whole debate. Academics were left arguing over noisy survey correlations in part because the cleanest causal data — a deactivation experiment run at platform scale — was generated privately, returned the wrong answer, and was shut down. “The science is mixed” turns out to describe the contents of the public literature, not the contents of Meta’s filing cabinet. We have run this experiment as a civilization once before, and a Meta employee named the precedent unprompted: tobacco. The published science on cigarettes stayed “contested” for exactly as long as the industry’s internal science stayed internal.
Staging the disease
Whatever the per-person effect size, the population-level readings are not subtle. American teens now spend about 40 minutes a day with friends in person, down from 140 minutes two decades ago, per the American Time Use Survey — face-to-face time replaced, hour for hour, by the substitute tissue. The US Surgeon General issued a formal advisory in 2023 noting that adolescents who spend more than three hours a day on social media face double the risk of anxiety and depression symptoms, then went further in 2024 and called for a tobacco-style warning label on the platforms themselves.
Read those as staging criteria. Within a single lifetime — within my lifetime, and these platforms are younger than I am — we have watched the presenting symptoms progress from “teens seem distracted” to the nation’s top public-health official asking Congress for cigarette labelling. That is not F1.
The first treatments are finally being prescribed
After twenty years of watchful waiting, actual interventions arrived in the last seven months — and they are blunter than anything the industry believed governments would dare.
Australia’s under-16 ban took effect on 10 December 2025: platforms from Instagram to TikTok to YouTube must take reasonable steps to prevent under-16s from holding accounts, on penalty of fines up to AU$49.5 million. Within weeks, platforms had removed 4.7 million under-age accounts. eSafety Commissioner Julie Inman Grant was precise about the goal: “success is measured by reduction in harm and in re-setting cultural norms.” Denmark is moving on an under-15 ban that could pass by mid-2026. Brazil’s Digital ECA took effect in March. Indonesia has ordered platforms closed to under-16s. More are queuing.
It is worth being precise about what these laws are, clinically. They are not a cure. Age bans treat exposure — they pull the most vulnerable patients away from the insult. The etiology is untouched: an advertising model that pays platforms in proportion to the attention they extract and the metadata they harvest will keep optimizing for the scar, for every user still on the feed. Nothing in any of these statutes changes what the machine is for.
And yet exposure control is exactly how you treat fibrosis. You cannot surgically remove scar tissue from a liver. What medicine does instead is remove the insult — stop the drinking, clear the hepatitis — because when the damage stops being inflicted, progression halts, and tissue sometimes partially regenerates. That is the hope buried inside the bleakest document in this whole record. Meta’s own suppressed study found that people felt measurably better after one week away. The buried evidence is simultaneously the indictment and the prognosis: the damage is real, the company knew, and the healing starts almost immediately once the insult is withdrawn.
Diseases do not get treated until they get named. Smoking had to become “lung cancer risk” before the warning labels came; leaded gasoline had to become “lead poisoning” before the phase-out. The platforms have spent twenty years making sure this disease stayed nameless — engagement, connection, community, time well spent. Call it what it is. The fabric is fibrotic, the biopsy results were on the table the whole time, and the only treatment ever shown to work is the one the patient has finally started demanding.