The percentage of Americans who think the government should "take steps to restrict false information online" is shrinking. - Pew also found decreasing support for the idea that tech companies should make such efforts on their own.
In 2023, 55 percent of people surveyed agreed with the idea that the U.S. government "should take steps to restrict false information online, even if it limits freedom of info." And 65 percent of those surveyed thought that tech companies should do this.
Support for both statements has decreased.
- In Pew's 2025 survey, just 51% agreed with the statement about the U.S. government, and
- 60% agreed with the statement about tech companies.
The
good news here is we may be seeing an uptick in people who view free
speech as a more important value than some utopian ideal of online
safety. Perhaps people have started to realize that systems tasked with
stopping false information online will inevitably make mistakes and
exhibit biases.
Speech that is categorically not misinformation
will sometimes get swept up, especially when companies are relying on
algorithms as a first line of misinformation defense.
Moreover,
what qualifies as misinformation is not always so simple to determine.
Things deemed misinformation at one point (like the lab leak hypothesis
of COVID-19 origins, or the Hunter Biden laptop story) may later turn
out to be true or at least possible. Things called misinformation by one
group of people may be seen differently by people with different
political, religious, or moral views. In many situations, saying who is
right and who isn't requires prioritizing one set of values over
another, or making judgement calls on complicated scientific matters
that may not be settled yet.
And even when some story or
piece of information is clearly false, suppressing it can cause more
damage than good. At the very least, there's no evidence that stopping
the spread of misinformation in certain online venues will actually stop
people from believing that information, especially when politicians and
offline pundits still spread it.
For those inclined to believe
the false story, its suppression could end up strengthening their
resolve that it is correct. Suppressing talk of the false story can even
deprive people of opportunities to counter it, with people directly
challenging the falsehoods often getting got up by the same suppression
rules as those spreading the falsehoods.
We've seen so many
examples of all of these scenarios in recent years that it would be
depressing if the message wasn't getting across somehow. So it's nice to
see a survey suggesting that, at least to some small degree, it is.
The Bad News
The
Pew survey still shows a lot of people who want the government to stop
misinformation online, and who are willing to sacrifice freedom of
information for this goal. For both the government and the tech
question, a majority are in support.
And while these percentages may have decreased since 2023, they're still higher than they were not very long ago.
In
2021, just 48 percent of people surveyed agreed with the idea that the
U.S. government should take step to restrict false information online
even when it required limiting freedom of in formations, and in 2018
just 39 percent of people agreed with this statement.
On
the question of tech companies suppressing false information, 59
percent agreed with that goal in 2021 and 56 percent agreed in 2018.
And
while Republican respondents answers on the government question haven't
varied too widely since 2018, Democratic responses seem to vary wildly
based on who is in power. In 2018, during Donald Trump's first term,
just 40 percent wanted the government to try to restrict false
information; in 2025, 58 percent wanted this. During Joe Biden's
presidency, some 65 percent (in 2021) and 70 percent (in 2023) thought
it was a good idea.
It's unsurprising that people will view their
side as more capable of good judgment on such matters. But the policies
and precedents set when their side is in power will survive far
afterward, making such lopsided judgments shortsighted. As long as Dems
believe censorship is OK when their people are in office, we'll continue
to see not only dangerous censorship under Democratic administrations
but no matter who is in power.
Suppressing Violent Content
Pew
also asked about support for government or tech companies suppressing
"violent content." Far fewer people in 2025 support either entity doing
so.
In 2023, 60 percent of those surveyed said the U.S.
government "should take steps to restrict extremely violent content
online, even if it limits freedom of info." In 2025, this was down to 52
percent.
As for tech companies suppressing violent content, support dropped from 71 percent in 2023 to 58 percent in 2025.
Subverting Democracy and Free Speech, for the Children
Who
needs to go through the whole messy democratic process of passing laws
when you can just declare major policy changes? That seems to be the
thought process of Missouri Attorney General Andrew Bailey. His office announced
this month that he has ordered adult websites "to verify the age of
users not just on the website but also on the device level—creating the
most robust age-verification standard in the country."
While
other states have debated such policies in their legislatures, Bailey is
trying to unilaterally impose this new policy under the auspice of his
consumer protection authority.
Bailey's new rule claims
that failing to check IDs or use some other form of "commercially
reasonable age verification technology" constitutes an "unfair practice"
under the Missouri Merchandising Practices Act (MMPA). Bailey also
claims that this is not a new policy at all, but merely a mechanism for
enforcing the state's existing prohibitions on distributing pornography
to people under age 18.
Even age verification rules adopted by
state legislatures keep getting blocked by courts, on the grounds that
they violate First Amendment rights. So it's possible that Bailey's
mandate here will suffer a similar fate. Regardless of what ultimately
happens, this attempt is notable because it represents an escalation of
attempts to age-gate the internet.
More Sex & Tech News
• Is the Department of Justice (DOJ) going to break up Google? Reason Foundation policy analyst Max Gulker breaks down
last week's verdict against the company in a case concerning online
advertising and antitrust law. Google "now must battle authorities eager
to force Google to sell off major business units in two major cases,"
writes Gulker:
Next week, the DOJ will argue in a
different court that Google should be forced to divest major business
units, potentially including its Chrome internet browser and Android
operating system, in the wake of an August 2024 guilty verdict for
antitrust violations related to online search.
Google has already
announced it will appeal Thursday's ad tech verdict. Pending appeals,
the case will proceed to a similar remedies phase, where the DOJ is
expected to seek the divestiture of a part of the ad tech business,
which generated 12% of parent company Alphabet Inc.'s revenue last year,
approximately $42 billion.
The ad tech case has been
overshadowed in the public debate by the search case and antitrust suits
against Amazon, Facebook, and Apple. The Trump Administration, usually
eager to reverse course on many of former President Joe Biden's
policies, has instead opted to continue all five major antitrust cases.
•
A blow for Trump's executive order on gender identity: "A federal judge
in Boston on Friday ordered the Trump administration to issue passports
that reflect the self-identified gender of six transgender people
rather than requiring that the passports display the sex on the
applicants' original birth certificates," The New York Times reports.
"The order from Judge Julia E. Kobick was a victory, at least
temporarily, for the six plaintiffs, who she said were likely to prevail
on their claim that a new policy by the Trump administration amounts to
a form of unconstitutional sex discrimination under the Fifth
Amendment, as well as the Administrative Procedures Act."
• ChatGPT developer OpenAI is considering launching its own social media platform. The Atlantic explains
the logic: "People create data every time they post online, and
generative-AI companies need a lot of data to train their products.
Social networks are also sticky: If you got hooked on an OpenAI feed,
you'd be less likely to use competing generative-AI products from
Anthropic or Google."
• Today in signs of backsliding: Some
states are urging the U.S. Supreme Court to reconsider its 2015 decision
legalizing same sex marriage. "No one is suggesting that
reconsideration of the decision in Obergefell is imminent," writes Amy Harmon in The New York Times.
"Still, the number of state measures proposed signals an effort to
shift the perception of same-sex marriage as an established civil right,
leaders on both sides of the issue say."
• "After a federal
court struck down their first attempt to limit teen social media use,
Ohio lawmakers have a new approach: go through the app stores," reports
Cleveland.com. "Under House Bill 226, app stores would ask users how
old they are. Users would verify their age and those younger than 16
would need parental permission to download apps on their phones." You
can find the full bill here.
• Customer service chatbots might "hallucinate" answers to support questions, as AI company Anysphere learned this week.
Immigration
Plus: A listener asks who was the better president: Trump or Obama?
Peter Suderman, Katherine Mangu-Ward, Nick Gillespie, and Matt Welch
Supreme Court
Understanding the Supreme Court's unusual late-night ruling against the Trump administration
Damo
