Meta relaxed its content moderation rules in the name of free speech. New research shows the consequences were swift and measurable: violent threats against Members of Congress on Facebook quadrupled in the six months that followed.
A report published June 9 by the Center for Countering Digital Hate (CCDH), titled “Safety Off,” analyzed nearly 8 million Facebook comments directed at 100 Members of Congress. It found that in the six months after Meta’s January 2025 policy overhaul, abusive comments tripled from approximately 24,000 to 78,000. Comments that violated Meta’s policies around violent threats specifically quadrupled, from 1,800 to 7,600. Threats against President Donald Trump more than doubled. Online abuse overall, including hate speech and harassment, surged across all metrics tracked.
The data points to a direct relationship between platform policy and real-world harm. The question the story raises goes beyond Meta: it exposes the fundamental difference between how the United States and Europe think about free speech, and how those philosophies produce radically different outcomes in how social media is governed.
Meta’s “More Speech and Fewer Mistakes” experiment
In January 2025, Meta overhauled its content moderation framework under the banner “More Speech and Fewer Mistakes.” Joel Kaplan, Meta’s chief global affairs officer, wrote at the time that the company had been “over-enforcing our rules, limiting legitimate political debate and censoring too much trivial content.”
The policy changes reduced automated enforcement, loosened restrictions on political speech, and scaled back the systems that flagged violent content, hate speech, and harassment. The stated goal was to default toward allowing speech rather than restricting it, a philosophy rooted in the American First Amendment tradition that the remedy for bad speech is more speech, not censorship.
CCDH’s analysis found that Meta’s own enforcement actions declined sharply in the same period. The report documented explicit calls for murder and suicide; racist, sexist, antisemitic, and homophobic abuse; and sexual harassment and doxxing attempts across the comments analyzed.
“When platforms stop enforcing their own rules against threats, hate, and harassment, they become complicit in normalizing intimidation and harassment of elected officials,” said Imran Ahmed, founder and CEO of CCDH. “Most people exposed to online hate will never commit violence, but it only takes one radicalized person to commit the most atrocious of crimes.”
The US approach: speech first, consequences later
The First Amendment to the US Constitution says Congress “shall make no law … abridging the freedom of speech.” This is an absolute prohibition on the government restricting expression. It creates a default where speech is protected unless it falls into a narrow set of exceptions: incitement to imminent violence, defamation, true threats, obscenity.
This philosophy shapes how US tech companies design their platforms. The default is to allow speech and only remove it after it causes harm. Section 230 of the Communications Decency Act shields platforms from liability for user content, meaning the legal and financial pressure to moderate is minimal. Meta’s policy shift was a natural expression of this approach: trust users, remove less, accept the consequences.
The US Supreme Court has consistently held that even offensive, hateful, or disturbing speech is constitutionally protected. In the landmark 1976 case Handyside v. United Kingdom, the European Court of Human Rights acknowledged this contrast directly, noting that European law covers even speech that may “offend, shock or disturb” but allows restrictions where necessary in a democratic society. The American approach offers no such flexibility.
The European approach: qualified rights, enforceable duties
Europe’s legal framework starts from a different premise. Article 10 of the European Convention on Human Rights guarantees freedom of expression, but immediately qualifies it. Paragraph 2 explicitly allows restrictions that are “prescribed by law and necessary in a democratic society” for purposes including national security, public safety, the prevention of disorder or crime, and the protection of the reputation or rights of others.
This is a qualified right, not an absolute one. The European Court of Human Rights applies a proportionality test, asking whether a restriction serves a legitimate aim and whether less restrictive alternatives exist. The “margin of appreciation” doctrine gives member states room to balance free expression against competing rights like privacy, dignity, and public order.
The practical difference is stark. Hate speech that is constitutionally protected in the US can be criminalized in Germany, France, or the UK. Holocaust denial is illegal in 18 European countries. The EU’s Digital Services Act requires platforms to proactively address illegal content, including threats and hate speech, or face significant fines.
Applied to Meta’s case, the European framework would have constrained the company’s ability to simply step back from moderation. The DSA’s systemic risk obligations require platforms of Facebook’s size to assess and mitigate risks to public security and civic discourse. A policy change that led to a quadrupling of violent threats would likely trigger regulatory action, not just critical headlines.
The competing visions
Meta’s experiment embodies the tension at the heart of the transatlantic debate. The US model trusts the marketplace of ideas to sort itself out, accepting that some harmful speech will circulate as the price of maximal freedom. The European model trusts democratic institutions to define boundaries, accepting that some legitimate speech may be restricted as the price of public safety.
Both approaches have vulnerabilities. The US model struggles to contain organized disinformation campaigns, targeted harassment, and the radicalization that can follow unchecked hate speech. The European model depends on state institutions wielding the restriction power responsibly, and critics point to authoritarian governments in Hungary and Poland (pre-2023) exploiting these carve-outs to suppress dissent.
Meta’s moderation rollback and its measurable consequences are not just a story about one company’s policy error. They are a real-world test of two competing philosophies of free expression, playing out on a platform with 3 billion users. The data from CCDH suggests that the American default of “more speech” has a measurable cost. Whether European voters and regulators consider that cost acceptable is a question that will shape the internet’s future.
Sources: WIRED (Jun 9, 2026); CCDH (Jun 9, 2026); Wikipedia — Article 10 of the European Convention on Human Rights; Wikipedia — First Amendment