Don’t Give Congress The Power To Censor The Internet — And Don’t Trust Social Media To Do It For Them

Caitlin Vogus / Jul 19, 2023

Caitlin Vogus is the Deputy Director of Advocacy at Freedom of the Press Foundation.

United States Senate website. Shutterstock

Some members of Congress have recently been loudly decrying what they see as government interference with online content moderation. Yet at the same time, some are now embracing a new proposal that would allow straight-up government censorship of the internet, at least when it benefits lawmakers’ own interests.

Senate Amendment 218 to the National Defense Authorization Act would turn social media and other online services into flunkies for federal lawmakers to scrub the internet of information about themselves. Amendment 218 makes it illegal to post or host information about lawmakers, their relatives and others on the internet if a member of Congress has demanded removal of that information.

Not only is Amendment 218 unconstitutional and opposed by a broad coalition of civil society organizations, it also foolishly trusts social media companies to moderate content well and to act in the best interests of users and the public. If passed, this proposal will make investigative journalism harder and lead to the removal of news and other important speech about lawmakers from the internet. Protecting lawmakers from actual threats to their safety is important, but giving Congress the power to censor the internet isn’t the right approach. Congress should not bundle legislation prioritizing its own privacy over the public interest into the must-pass defense bill.

Here’s how the law would work: A lawmaker, their family member, or even anyone just living in the same household as them makes a written demand to a person, business, or association to remove certain information publicly posted about them online. The information could include things like their address, email address, license plate number, or the identity of their child. The person or company notified must then remove it within 72 hours and ensure the information isn’t made available on any website they control. If the person or company doesn’t comply, the lawmaker can sue.

The problem? This is a classic prior restraint, forbidden by the First Amendment in all but the most exceptional circumstances. However, even if Amendment 218 weren’t unconstitutional, it would still be a bad idea. Even if it sounds at first blush like a reasonable safety protection for members of Congress, it would be disastrous for online speech that has nothing to do with threats to lawmakers’ security.

Just consider how much news reporting and public oversight rely on access to precisely the kind of information that Amendment 218 would delete from the internet. For example, a journalist reporting about a lawmaker’s vacation in the midst of a natural disaster in his state may need to visit the legislator’s home to verify he isn’t there. A news outlet reporting on a congressman accused of using a sham car sale to cover up a bribe may need to verify his vehicle purchases and ownership. A reporter investigating a representative accused of using his staff for personal errands, such chauffeuring his children, may need to know who his children are and how they get to school. These important news stories may never get written if the information journalists need to track down the facts isn’t available online.

Amendment 218 will also inevitably lead to the removal of news reports from social media, despite an exception for information that is part of a news story or other speech on matters of public concern. Both anecdotal evidence and actual research show that the automated tools social media companies use for content moderation are notoriously bad at accurately understanding what users are actually saying online. Companies will inevitably mistake journalism and other legitimate and valuable speech about members of Congress for “prohibited” posts and remove them.

And even if social media companies could accurately understand the content of every post for which they receive a notification, we can’t trust them to judge whether a particular post is about a matter of public concern — a necessary precondition to qualify for the exception — or, for that matter, to care. A company that receives a notification claiming that a post threatens legislators’ personal safety is likely to just remove it without considering the potential public interest in the post. Companies won’t want to displease the very members of Congress who have the power to regulate them.

In addition, the pressure to remove content — no questions asked — is heightened when companies have just 72 hours to comply with removal notifications. This short turnaround time doesn’t exactly encourage social media companies to carefully weigh and deliberate what constitutes news and whether a particular post is about a matter of public concern.

But even if a company cares enough to evaluate the post to see if the exception applies, the amendment contains no guidelines for determining whether speech is about a matter of public concern. Social media companies are ill-equipped to make that decision on their own. Companies’ own biases about what’s important to the public will inevitably influence their choice about whether to remove a post or leave it up under the exception. The posts that Truth Social or Facebook, for instance, consider to be about matters of public concern are likely to be very different.

In rare instances, a company may refuse to remove a post because it decides it is about a matter of public concern. But in that case, lawmakers can sue, leaving users reliant on companies to pursue a legal fight rather than cave and remove the post. Amendment 218 itself doesn’t give users any right to defend their speech in court. A user may be able to bring a separate First Amendment lawsuit against the legislator who initiated the removal of their speech. But even if the user wins that suit, there’s no mechanism to require the social media company to put the post back up, and no guarantee that a company will.

Finally, not only does Amendment 218 require an online company to remove a specific instance of covered information, but it must ensure that information does not appear elsewhere on any of its websites. This “stay down” requirement goes beyond a simple takedown of any one instance of information appearing on a website. It will encourage companies to attempt to screen and filter user-generated content across their services to ensure prohibited information about a member of Congress doesn’t appear again, inevitably resulting in even more overremovals of online speech.

In short, this proposal will result in a huge loss of access to news and information about Congress. It will give senators and representatives a censor’s pen to blot out information that they find embarrassing, inconvenient, or politically damaging. The Senate shouldn’t empower itself to control online speech. It should reject Amendment 218.


Caitlin Vogus
Caitlin Vogus is the Deputy Director of Advocacy at Freedom of the Press Foundation, where she works to defend and protect press freedoms, journalists, and whistleblowers.