A new report shows that violent threats against lawmakers have surged online after Big Tech companies such as Meta have rolled back moderation policies and program algorithms to prioritize inflammatory content.
Now, Utah Sen. John Curtis wants to do something about it.
“People have the right to say what they want to say,” Curtis told the Deseret News in a statement. “People can say whatever they want, that’s not the problem. The problem is when the social media companies use their algorithms to amplify those, and so they get to decide, ‘OK, this is a message we want 100 million people to see, we’re just going to cram it down their throats.’”
A recent report from the Center for Countering Digital Hate found that violent threats online have quadrupled since Meta implemented a new policy in January 2025 reining in its moderation practices. Since then, instances of threats, hate and harassment all tripled in that time frame.
The study analyzed nearly 8 million comments on Facebook directed toward 100 members of Congress over a yearlong period — six months before the moderation policy change and six months after. The report found that abusive comments rose from roughly 24,000 comments before the policy change to 78,000 comments afterward.
Violent threats quadrupled from 1,800 comments to 7,600 comments; harassment rose from 15,700 comments to nearly 40,000; and threats directly targeting President Donald Trump doubled in that time period.
But the problem, Curtis argues, isn’t necessarily the comments themselves. Instead, the fault lies with social media companies who amplify hateful comments to increase engagement — thereby creating a reward system for bad behavior.
“What these social media companies are doing, is they’re ginning up this anger by selecting some of the worst messages, amplifying them, and there’s no accountability for that,” Curtis said. “Section 230 protects them from that. So I feel that this is a little bit of — in legal terms, they call it duty of care. We need to hold them responsible for that, and I think if you did that one thing alone, this would dramatically change.”
Curtis has long called for reforms to the social media landscape, even going so far as to introduce a bill late last year proposing updates to Section 230 to no longer offer protections for tech companies if it can be proved that the platforms knowingly used an algorithm to amplify content responsible for causing harm. The change would instead mandate that companies “own” the content, which would allow individuals to sue them directly.
That bill, the Algorithm Accountability Act, has yet to get a vote in Congress.
In practice, the bill would establish requirements for social media platforms to “design, train, test, deploy, operate, and maintain their recommendation-based algorithm to prevent foreseeable bodily injury or death.” It would also provide a “clear civil right of action” to take legal action.
The debate comes as lawmakers have long grappled with whether content moderation on social media platforms amounts to a suppression of speech. But while Republicans and Democrats have often found themselves on the opposite side of that debate, Curtis said guardrails can be put in place by simply not amplifying hateful content without stripping anyone of their right to express an opinion.
“If you watch the hearings between Republicans and Democrats, you’ll get comments like Republicans saying ‘Why did you censor this,’ and the Democrats saying, ‘Why didn’t you censor this,’” Curtis said, “That’s a debate I don’t think anybody’s going to win about what can be said and what can’t be said. But I do believe you can, you can solve that quickly by just not amplifying the things that are said, and then they kind of die on their own.”