Last winter, my mother called me in a panic. A video in her Facebook feed claimed that common blood pressure medication had been quietly recalled because it caused cancer. She had already stopped taking it. The video was false. It had been shared thousands of times before anyone flagged it, and by then my mother had skipped four days of a medication her doctor considered essential.

I called the pharmacy, then her doctor, then sat with her while she took the pill again. Afterward, I went looking for who was responsible. The answer, under American law, is almost nobody. Section 230 of the Communications Decency Act says that no provider of an interactive computer service "shall be treated as the publisher or speaker" of information provided by someone else. Those 26 words, written when most Americans had never been online, have become the most powerful legal shield in the technology industry.

The law made sense in its time. Congress wanted to protect fledgling message boards from being sued over every comment a user posted. Nobody in 1996 imagined

companies worth more than a trillion dollars using machine-learning systems to decide which posts billions of people see first.

That is the heart of the problem. Today's social media platforms are not passive bulletin boards. They rank, recommend, promote and monetize. When a lie goes viral, it rarely does by accident.

Social media is now a primary news source for tens of millions of Americans. According to the Pew Research Center, 53 percent of U.S. adults get news from social media at least sometimes. Facebook is a regular news source for 38 percent of adults, YouTube for 35 percent, and TikTok and Instagram for 20 percent each. Among TikTok users, the share who regularly get news there has climbed from 22 percent in 2020 to 55 percent in 2025.

Falsehood travels faster than truth on these platforms. A landmark 2018 study by MIT researchers, published in the journal Science, analyzed roughly 126,000 stories on Twitter and found that false news was 70 percent more likely to be retweeted than true news, reaching 1,500 people about six times faster. Recommendation engines reward what people react to, and outrage and shock produce the strongest reactions.

Meanwhile, the companies have retreated from the job. In January 2025, Meta ended its third-party fact-checking program in the United States and replaced it with crowd-sourced Community Notes. Mark Zuckerberg said at the time that fact-checkers had been "too politically biased." Whatever one thinks of that claim, the result is a company with more than three billion users relying largely on volunteers to correct falsehoods.

How many people saw the false claim before the note appeared? How many never saw the note?

For years, judges read Section 230 broadly. That consensus is cracking.

In 2024, the Third Circuit Court of Appeals ruled in Anderson v. TikTok that the platform could face liability after its algorithm promoted the deadly "Blackout Challenge" to a 10-year-old girl. The court reasoned that algorithmic curation is the company's own expression, not merely third-party speech. Other courts disagree. The Ninth Circuit declined to follow Anderson this year in a case over Facebook's role in violence against the Rohingya in Myanmar, although two concurring judges said the Third Circuit got it right. Legal analysts now describe a clear circuit split that the Supreme Court may eventually have to resolve.

In March, a Los Angeles jury found Meta and YouTube negligent in the first social media addiction case to reach trial, awarding $6 million to a young woman who began using the platforms as a child. Villanova law professor Peter Ormerod called the verdict "a momentous development." Cornell's Sarah Kreps noted there are "thousands pending." That case concerned addictive design rather than misinformation. The principle is the same: when a company designs a product that causes foreseeable harm, it should answer for it.

Congress is stirring too. Seven senators have backed the Sunset Section 230 Act, and a bipartisan House bill introduced last month would end the protections two years after enactment.

Defenders of Section 230 raise serious concerns. The Electronic Frontier Foundation calls the law "the legal support that sustains the internet as we know it." Without it, they argue, platforms would either censor aggressively to avoid lawsuits or stop moderating altogether. Smaller sites, without armies of lawyers, could be sued out of existence, leaving the giants even more dominant.

There is also the hardest question of all: who decides what counts as misinformation? Health guidance changes. Political claims are contested. Some ideas once dismissed as fringe later proved credible. A law that punishes platforms for hosting "false" speech could quickly become a weapon for whichever party holds power. The First Amendment protects a great deal of speech that is simply wrong.

These are real dangers. They argue for careful reform, not for leaving the shield untouched.

Nobody sensible wants Facebook sued over every false post written by a stranger. The target should be narrower: the platform's own conduct.

First, Congress should clarify that Section 230 does not protect algorithmic amplification. A company that chooses to push a false medical claim into millions of feeds because it drives engagement is no longer a neutral host. Second, large platforms should face transparency requirements, including independent audits of how their recommendation systems treat known falsehoods. The European Union already demands this under its Digital Services Act, which allows fines of up to 6 percent of global annual revenue. Third, platforms should give independent researchers meaningful access to data, offering the public an alternative to company press releases. Fourth, any reform should protect small companies and offer clear safe harbors for good-faith content moderation.

Would this end up misinformation? No. Lies predate the internet. It would end the pretense that companies earning billions from engagement bear no responsibility for what that engagement rewards.

My mother is fine now. She checks with her doctor before believing anything she reads online. Millions of Americans do not have a pushy adult child on the other end of the phone.

Next time a shocking claim appears in your feed, ask yourself who decided you should see it, and why. If the answer is an algorithm built to maximize profit, shouldn't the company that built it share the blame when that claim turns out to be false?

If Algorithms Profit From Lies, Why Should Platforms Escape All the Blame?

Similar Opinions

Latest in The Opinion Times

Are Cheap Chinese EVs Dismantling Europe's Car Industry?

Trump Promised Young Voters Relief; Why Are Three in Four Now Unconvinced?

Will AI Build Tomorrow’s Careers, or Quietly Close the Door on Graduates?

If Algorithms Profit From Lies, Why Should Platforms Escape All the Blame?

From Boats to Jungle Bases, Without a Single Congressional Vote