Originally published 2020-05-29.
On 28 May 2020, President Donald Trump signed Executive Order 13925, "Preventing Online Censorship", which aimed to weaken the legal shield that protects platforms such as Twitter, Facebook and Google from lawsuits over what their users post. He signed it two days after Twitter, for the first time, attached a fact-check label to one of his tweets. In practice the order changed very little: the regulators it relied on never acted, and President Joe Biden revoked it in May 2021.
The fight behind it, over Section 230 of the Communications Decency Act, has not gone away. Here is what the order tried to do, why it had limited legal force, and what the law it targeted actually says.
What sparked the order
On 26 May 2020 Trump posted a series of tweets claiming that mail-in ballots would be "substantially fraudulent" and singling out California, which was planning to send ballots to registered voters because of the pandemic.
There is NO WAY (ZERO!) that Mail-In Ballots will be anything less than substantially fraudulent. Mail boxes will be robbed, ballots will be forged & even illegally printed out & fraudulently signed. The Governor of California is sending Ballots to millions of people, anyone.....
— Donald J. Trump (@realDonaldTrump) May 26, 2020
Twitter did not delete the tweets. It added a small label inviting readers to "get the facts" about mail-in voting, linking to news reports and context. The label was applied under the company's policy on misleading election information by its site integrity team, headed by Yoel Roth. It was the first time Twitter had taken any action against a Trump tweet.
Trump responded the next day by accusing social media companies of silencing conservative voices and threatening to regulate them strongly or shut them down.
Republicans feel that Social Media Platforms totally silence conservatives voices. We will strongly regulate, or close them down, before we can ever allow this to happen. We saw what they attempted to do, and failed, in 2016. We can’t let a more sophisticated version of that....
— Donald J. Trump (@realDonaldTrump) May 27, 2020
He also attacked Roth by name over posts Roth had written before the label, calling him a "hater". Roth later described receiving a wave of harassment.
So ridiculous to see Twitter trying to make the case that Mail-In Ballots are not subject to FRAUD. How stupid, there are examples, & cases, all over the place. Our election process will become badly tainted & a laughingstock all over the World. Tell that to your hater @yoyoel
— Donald J. Trump (@realDonaldTrump) May 28, 2020
What the executive order directed
The order argued that large platforms were engaging in "selective censorship" and therefore should not enjoy Section 230's full protection when they edit or restrict content in ways that are not in "good faith". Speaking at the signing, Trump said of the companies: They have a shield. They can do what they want, they're not going to have that shield.
Its main instructions were:
- The National Telecommunications and Information Administration (NTIA), part of the Commerce Department, had 60 days to petition the Federal Communications Commission (FCC) to issue rules clarifying when moderation is in good faith.
- The Federal Trade Commission was encouraged to consider action against platforms whose moderation did not match their public promises, and the White House would collect complaints of online political bias.
- Federal agencies were told to review their advertising spending on platforms accused of restricting speech.
#BREAKING: President Trump signs executive order strip liability protection from companies that censure content: "Companies that engage in censoring or any political conduct will not be able to keep their liability shield." https://t.co/D5ooUw1fNz pic.twitter.com/FHs7kUvJH1
— The Hill (@thehill) May 28, 2020
Section 230 in plain English
Section 230 was passed in 1996 as part of the Communications Decency Act, itself Title V of the Telecommunications Act. Its two key provisions are short.
The first, section 230(c)(1), says an online service is not to be treated as the publisher or speaker of information provided by someone else. So if a user posts something defamatory, the user can be sued, but the platform generally cannot.
The second, section 230(c)(2), protects services when they remove or restrict material they consider obscene, violent, harassing or "otherwise objectionable", provided they act in good faith. This is the clause the executive order tried to narrow.
The law was written by Representatives Chris Cox and Ron Wyden partly in response to a 1995 court case, Stratton Oakmont v. Prodigy, which held that a service that moderated its boards could be liable as a publisher, while an earlier case had cleared CompuServe because it did not moderate at all. Lawmakers did not want to punish companies for trying to clean up their sites. In 1997, Zeran v. AOL set the broad reading of the immunity that courts have followed since.
Why the order had limited effect
An executive order cannot rewrite a law passed by Congress. Only Congress can amend Section 230, and only courts can decide how it applies in particular cases. The order could only ask agencies to reinterpret it.
That strategy depended on the FCC, an independent agency that is not bound to follow presidential instructions. The NTIA did file its petition, but the FCC never acted on it. A civil liberties group also sued, arguing that the order punished Twitter for its own speech in labelling the president's tweets.
In May 2021, President Biden rescinded Executive Order 13925 along with several other Trump orders, formally ending the effort.
Common misunderstandings
"Twitter censored the president." In May 2020 it did not remove or hide his tweets; it added a link to more information. Then-CEO Jack Dorsey said the aim was to connect the dots of conflicting statements
so people could judge for themselves.
"Platforms must be neutral to keep Section 230." The law contains no neutrality requirement. The idea that a site must choose between being a "platform" and a "publisher" is a popular myth; the statute protects moderation rather than forbidding it.
"Section 230 protects anything online." It does not cover federal criminal law or intellectual property, and in 2018 the FOSTA-SESTA law removed protection for content that knowingly facilitates sex trafficking.
What happened next
The clash between Trump and the platforms escalated. After the attack on the US Capitol in January 2021, Twitter suspended his account permanently and Facebook suspended him too. Trump went on to launch his own platform, Truth Social, and his Twitter account was restored in November 2022 after Elon Musk bought the company.
Section 230 itself survived. Politicians on both the right and the left have proposed changing it, for very different reasons, but no broad reform has passed. In 2023 the Supreme Court heard Gonzalez v. Google, a case widely expected to test the law's limits, but decided it on other grounds and left Section 230 untouched.
The 2020 order is best remembered as an early marker in that long argument: a high-profile attempt to change how social media is moderated, made without the legal power to do it.
Sources: Next TV: President Biden axes Trump attack on social media (17 May 2021); Wikipedia: Section 230; Wikipedia: Yoel Roth; contemporary reports by the Los Angeles Times, The Verge, Politico and the Financial Times (28 May 2020).
COMMENTS