U.S.

DEPARTMENT OF JUSTICE

Section 230 — Nurturing Innovation

or Fostering Unaccountability?

KEY TAKEAWAYS AND RECOMMENDATIONS

June 2020
 UNITED STATES DEPARTMENT OF JUSTICE’S
REVIEW OF SECTION 230
OF THE COMMUNICATIONS DECENCY ACT OF 1996

As part of its broader review of market-leading online platforms, the U.S. Department of
Justice analyzed Section 230 of the Communications Decency Act of 1996, which provides immunity
to online platforms from civil liability based on third-party content and for the removal of content
in certain circumstances. Congress originally enacted the statute to nurture a nascent industry
while also incentivizing online platforms to remove content harmful to children. The combination of
significant technological changes since 1996 and the expansive interpretation that courts have given
Section 230, however, has left online platforms both immune for a wide array of illicit activity on their
services and free to moderate content with little transparency or accountability.

The Department of Justice has concluded that the time is ripe to realign the scope of Section
230 with the realities of the modern internet. Reform is important now more than ever. Every year,
more citizens—including young children—are relying on the internet for everyday activities, while
online criminal activity continues to grow. We must ensure that the internet is both an open and
safe space for our society. Based on engagement with experts, industry, thought leaders, lawmakers,
and the public, the Department has identified a set of concrete reform proposals to provide stronger
incentives for online platforms to address illicit material on their services, while continuing to foster
innovation and free speech.
***
The Department's review of Section 230 arose in the context of our broader review of market-
leading online platforms and their practices, announced in July 2019. While competition has been
a core part of the Department’s review, we also recognize that not all concerns raised about online
platforms (including internet-based businesses and social media platforms) fall squarely within the
U.S. antitrust laws. Our review has therefore looked broadly at other legal and policy frameworks
applicable to online platforms. One key part of that legal landscape is Section 230, which provides
immunity to online platforms from civil liability based on third-party content as well as immunity for
removal of content in certain circumstances.1

Drafted in the early years of internet commerce, Section 230 was enacted in response to a
problem that incipient online platforms were facing. In the years leading up to Section 230, courts
had held that an online platform that passively hosted third-party content was not liable as a publisher
if any of that content was defamatory,2 but that a platform would be liable as a publisher for all its

1
The two key operative provisions of Section 230 are: (1) “No provider or user of an interactive computer service shall
be treated as the publisher or speaker of any information provided by another information content provider,” and (2) “No
provider or user of an interactive computer service shall be held liable on account of…any action voluntarily taken in good
faith to restrict access to or availability of material that the provider or user considers to be obscene, lewd, lascivious,
filthy, excessively violent, harassing, or otherwise objectionable, whether or not such material is constitutionally
protected,” 47 U.S.C. § 230(c)(1), (2).
2
Cubby, Inc. v. CompuServe, Inc., 776 F. Supp. 135 (S.D.N.Y 1991).

1
third-party content if it exercised discretion to remove any third-party material.3 Platforms therefore
faced a dilemma: They could try to moderate third-party content but risk being held liable for any and
all content posted by third parties, or choose not to moderate content to avoid liability but risk having
their services overrun with obscene or unlawful content. Congress enacted Section 230 in part to
resolve this quandary by providing immunity to online platforms both for third-party content on their
services or for removal of certain categories of content. The statute was meant to nurture emerging
internet businesses while also incentivizing them to regulate harmful online content.

The internet has changed dramatically in the 25 years since Section 230’s enactment in ways
that no one, including the drafters of Section 230, could have predicted. Several online platforms
have transformed into some of the nation’s largest and most valuable companies, and today’s online
services bear little resemblance to the rudimentary offerings in 1996. Platforms no longer function as
simple forums for posting third-party content, but instead use sophisticated algorithms to promote
content and connect users. Platforms also now offer an ever-expanding array of services, playing an
increasingly essential role in how Americans communicate, access media, engage in commerce, and
generally carry on their everyday lives.

These developments have brought enormous benefits to society. But they have also had
downsides. Criminals and other wrongdoers are increasingly turning to online platforms to engage
in a host of unlawful activities, including child sexual exploitation, selling illicit drugs, cyberstalking,
human trafficking, and terrorism. At the same time, courts have interpreted the scope of Section 230
immunity very broadly, diverging from its original purpose. This expansive statutory interpretation,
combined with technological developments, has reduced the incentives of online platforms to address
illicit activity on their services and, at the same time, left them free to moderate lawful content
without transparency or accountability. The time has therefore come to realign the scope of Section
230 with the realities of the modern internet so that it continues to foster innovation and free speech
but also provides stronger incentives for online platforms to address illicit material on their services.

Much of the modern debate over Section 230 has been at opposite ends of the spectrum.
Many have called for an outright repeal of the statute in light of the changed technological landscape
and growing online harms. Others, meanwhile, have insisted that Section 230 be left alone
and claimed that any reform will crumble the tech industry. Based on our analysis and external
engagement, the Department believes there is productive middle ground and has identified a set of
measured, yet concrete proposals that address many of the concerns raised about Section 230.

A reassessment of America’s laws governing the internet could not be timelier. Citizens are
relying on the internet more than ever for commerce, entertainment, education, employment, and
public discourse. School closings in light of the COVID-19 pandemic mean that children are spending
more time online, at times unsupervised, while more and more criminal activity is moving online. All
of these factors make it imperative that we maintain the internet as an open and safe space.

3
Stratton Oakmont, Inc. v. Prodigy Servs. Co., 1995 WL 323710 (N.Y. Sup. Ct. 1995).

2
The Section 230 reforms that the Department of Justice identified generally fall into four categories:

1) Incentivizing Online Platforms to Address Illicit Content

The first category of potential reforms is aimed at incentivizing platforms to address the
growing amount of illicit content online, while preserving the core of Section 230’s immunity for
defamation.

a. Bad Samaritan Carve-Out. First, the Department proposes denying Section 230
immunity to truly bad actors. The title of Section 230’s immunity provision—“Protection
for ‘Good Samaritan’ Blocking and Screening of Offensive Material”—makes clear that
Section 230 immunity is meant to incentivize and protect responsible online platforms.
It therefore makes little sense to immunize from civil liability an online platform that
purposefully facilitates or solicits third-party content or activity that would violate federal
criminal law.

b. Carve-Outs for Child Abuse, Terrorism, and Cyber-Stalking. Second, the Department
proposes exempting from immunity specific categories of claims that address particularly
egregious content, including (1) child exploitation and sexual abuse, (2) terrorism, and (3)
cyber-stalking. These targeted carve-outs would halt the over-expansion of Section 230
immunity and enable victims to seek civil redress in causes of action far afield from the
original purpose of the statute.

c. Case-Specific Carve-Outs for Actual Knowledge or Court Judgments. Third, the

Department supports reforms to make clear that Section 230 immunity does not apply in a
specific case where a platform had actual knowledge or notice that the third party content
at issue violated federal criminal law or where the platform was provided with a court
judgment that content is unlawful in any respect.

2) Clarifying Federal Government Civil Enforcement Capabilities

A second category of reform would increase the ability of the government to protect citizens
from illicit online conduct and activity by making clear that the immunity provided by Section 230
does not apply to civil enforcement by the federal government, which is an important complement to
criminal prosecution.

3
 3) Promoting Competition
A third reform proposal is to clarify that federal antitrust claims are not covered by Section
230 immunity. Over time, the avenues for engaging in both online commerce and speech have
concentrated in the hands of a few key players. It makes little sense to enable large online platforms
(particularly dominant ones) to invoke Section 230 immunity in antitrust cases, where liability is based
on harm to competition, not on third-party speech.

4) Promoting Open Discourse and Greater Transparency

A fourth category of potential reforms is intended to clarify the text and original purpose of
the statute in order to promote free and open discourse online and encourage greater transparency
between platforms and users.

a. Replace Vague Terminology in (c)(2). First, the Department supports replacing the vague
catch-all “otherwise objectionable” language in Section 230 (c)(2) with “unlawful” and
“promotes terrorism.” This reform would focus the broad blanket immunity for content
moderation decisions on the core objective of Section 230—to reduce online content
harmful to children—while limiting a platform's ability to remove content arbitrarily or in
ways inconsistent with its terms or service simply by deeming it “objectionable.”

b. Provide Definition of Good Faith. Second, the Department proposes adding a statutory
definition of “good faith,” which would limit immunity for content moderation decisions
to those done in accordance with plain and particular terms of service and accompanied
by a reasonable explanation, unless such notice would impede law enforcement or risk
imminent harm to others. Clarifying the meaning of "good faith" should encourage
platforms to be more transparent and accountable to their users, rather than hide behind
blanket Section 230 protections.

c. Continue to Overrule Stratton Oakmont to Avoid the Moderator’s Dilemma. Third, the
Department proposes clarifying that a platform’s removal of content pursuant to Section
230 (c)(2) or consistent with its terms of service does not, on its own, render the platform
a publisher or speaker for all other content on its service.

* * *

4
 These reforms to Section 230 immunity will incentivize online platforms to police responsibly
content that is illegal and exploitive while continuing to encourage a vibrant, open, and competitive
internet. These twin objectives of giving online platforms the freedom to grow and innovate while
also encouraging them to moderate obscene and unlawful content were the core objectives of Section
230 at the outset. The Department’s proposed reforms aim to realize these objectives more fully and
clearly so that, in light of the vast technological changes since 1996, Section 230 better serves the
interests of the American people.

5
 OVERVIEW OF DEPARTMENT OF JUSTICE'S REVIEW OF SECTION 230

The Department of Justice’s review of Section 230 of the Communications Decency Act
included a number of different components:

1. Public Workshop. On February 19, 2020, the Department held a public workshop on
Section 230 titled “Section 230 — Nurturing Innovation or Fostering Unaccountability?”
The workshop gathered experts with diverse views on the benefits and drawbacks of
Section 230 and potential reforms. The Attorney General and FBI Director gave opening
remarks, followed by three panel discussions with a variety of thought leaders on Section
230, including litigators, academics, victims’ representatives, industry representatives, and
other experts. The Department of Justice livestreamed the event, https://www.justice.
gov/opa/video/section-230-workshop-nurturing-innovation-or-fostering-unaccountability,
and drafted a written summary.

2. Expert Roundtable. In the afternoon of February 19, 2020, following the public workshop,
the Department hosted a roundtable discussion with additional thought leaders
representing diverse viewpoints to discuss Section 230 and potential reforms in further
detail. To facilitate a robust dialogue, the afternoon session operated under the Chatham
House Rule. The Department drafted a written summary of the roundtable discussion to
synthesize the topics discussed, while anonymizing the contributions of the speakers.

3. Written Submissions. Participants in the morning Workshop and afternoon Roundtable

were also invited to submit short written statements with their views on Section 230,
which the Department reviewed.

4. Industry Listening Sessions. Following the Workshop, the Department met individually
with companies that had attended the public event or otherwise expressed interest in
discussing Section 230. The companies included internet and traditional firms, with
diverse businesses and different perspectives on the benefits and harms of Section 230.
Meetings were private and confidential to foster frank discussions about how these firms
employ Section 230 and their views on potential changes.

The Section 230 Workshop Agenda, Biographies of Participants, Workshop and Roundtable
Summary, and Participant Written Submissions are all available at https://justice.gov/ag/department-
justice-s-review-section-230-communications-decency-act-1996.

7
 DEPARTMENT OF JUSTICE KEY TAKEAWAYS AND
AREAS RIPE FOR SECTION 230 REFORM

Based on what we learned from this process, the Department drafted a set of key takeaways,

9
 PART I: KEY PRINCIPLES ON SECTION 230 REFORM

1. Recognize that Large Tech Platforms Are No Longer Nascent or Fragile. Since the enactment
of Section 230 almost 25 years ago, the internet and social media ecosystems have grown
exponentially. So too have the leading internet and social media companies, which today
are some of the most valuable American enterprises. The transformation of a few start-up
internet companies into vast transnational enterprises, boasting annual revenues exceeding
that of many countries, has raised valid questions of whether those large tech companies still
require the blanket immunity Section 230 provided to the nascent internet industry. While
it may be imprudent to repeal the immunity entirely, it seems clear that tailored changes to
immunity to make the internet a safer place would not unduly burden large tech companies.

2. Preserve Competition. Section 230 immunity has facilitated the growth of newer tech
startups, which have fewer resources to police content and to defend lawsuits than the large
tech platforms. At the same time, many are concerned that large online platforms are able
to use Section 230 immunity to maintain their dominant positions and avoid regulations or
responsibilities that apply to offline competitors, effectively distorting competition. Reforms
to Section 230 should aim to preserve and promote competition. Reforms should avoid
imposing significant compliance costs on small firms that could raise barriers to entry and
entrench dominant platforms and, where possible, should help level the playing field.

3. Keep Core Immunity for Defamation To Foster Free Speech. One of the key benefits and
most frequent uses of Section 230 identified by experts and industry is the protection for
online platforms from intermediary liability for defamation based on material posted by third
parties on their services. Such immunity is important to avoid chilling free speech online
and to promote pro-consumer business models, such as hosting user-generated content
and consumer reviews. Platforms generally do not have cost-effective means of evaluating
whether a third-party statement is defamatory because they typically do not know if the third-
party statement is true or false when made. For example, a review platform generally cannot
tell if the statement from a restaurant patron that “the soup was cold” is true or false. Absent
immunity, platforms that choose to permit online consumer reviews and third-party content
would be likely to remove or censor content whenever someone objects in order to minimize
liability risks. Removing the core immunity against defamation claims could imperil certain
online business models, particularly for smaller companies. It also could give a heckler’s veto
to anyone who objects to third-party content, regardless of whether the objection has merit.

11
4. Distinguish between Hosting Defamatory Content and Enabling Criminal Activity. While the
Department's Section 230 Workshop participants expressed concern about over-censorship if
immunity were removed for defamation claims, some participants emphasized the distinction
between immunity for defamation-type claims and immunity for enabling dangerous or
criminal behavior. Several experts suggested, and the Department generally agrees, that the
benefits of immunity do not outweigh the costs when it comes to enabling serious offenses
and harms, such as child sexual abuse material and terrorism-related offenses. In such cases,
the balance weighs in favor of protecting individuals from harm and giving victims meaningful
redress. Where possible, reforms to Section 230 should distinguish between the core
immunity for defamation-type torts and claims based on third-party content that facilitates or
constitutes federal criminal activity.

12
 PART II: ISSUES RIPE FOR LEGISLATIVE REFORM

Reforms to Better Incentivize Online Platforms to Address Illicit Content

1. Exempting Bad Actors from Blanket Section 230 Immunity

Carve-Out for Bad Actors Who Purposefully Facilitate Criminal Activity or Material

Throughout the Department’s discussions, there seemed to be general agreement that

online platforms that purposefully promotes, solicits, or facilitates criminal activity by third parties
should not receive the benefit of Section 230 immunity, an immunity described in the statute as
“Good Samaritan” immunity. A Good Samaritan is someone who stops to help a stranger in need, not
someone who endangers the stranger in the first place. To reward bad actors with blanket Section 230
immunity is inconsistent with the purpose of the statute.

Under the current expansive interpretation of Section 230, even websites designed to
promote or facilitate illegal conduct can still enjoy the protections of Section 230 immunity. See,
e.g., Daniel v. Armslist, LLC, 926 N.W.2d 710, 715, 726 (Wis. 2019), cert. denied. 140 S.Ct. 562 (2019)
(website that facilitated sale of firearm to prohibited person who then murdered wife and two other
people, and injured four others was immune under Section 230, despite allegations that website was
intentionally designed with the specific purpose of skirting federal firearm laws); Jones v. Dirty World
Entm't Recordings LLC, 755 F.3d 398, 406 (6th Cir. 2014) (reversing the district court’s withholding of
Section 230 immunity for “a website owner who intentionally encourages illegal or actionable third-
party postings to which he adds his own comments ratifying or adopting the posts”).

The Department views an explicit “Bad Samaritan” Carve-Out as necessary to ensure that bad
actors do not benefit from Section 230’s sweeping immunity at the expense of their victims. If, for
example, an online platform purposefully solicits third parties to sell illegal drugs to minors, exchange
child sexual abuse material, or otherwise engage in criminal activity on its service, it should not have
the ability to turn around and assert blanket Section 230 immunity in all private civil cases.

While experts and industry widely accepted the notion of separating the truly bad actors from
the rest, some expressed concerns about a broadly-worded provision that could expose good actors to
frivolous litigation. We believe that a few cabining principles would help alleviate this concern.

First, a “Bad Samaritan” Carve-Out should have a heightened mens rea, such as “purposefully,”
under which platforms that accidently or even negligently facilitate unlawful behavior would not lose
immunity. Requiring that a platform act purposefully or knowingly would also not impose a burden on
platforms to proactively screen all third-party content where there is no indication that the platform

14
is being used to engage in unlawful activity. At the same time, platforms should not receive immunity
where they purposefully promote, solicit, or facilitate the posting of material that the platform knew
or had reason to believe would violate federal criminal law.

A “Bad Samaritan” Carve-Out could be enacted instead of, or in addition to, proposals for
platforms to earn Section 230 immunity. They are not mutually exclusive. If enacted together with
such a proposal, a “Bad Samaritan” Carve-Out would provide a safety net or statutory guardrail on top
of a government or private standard-setting process.

Carve-Out for Actors Who Purposefully Blind Themselves and Law Enforcement to Illicit Material

As with the “Bad Samaritan” Carve-Out, it makes little sense to apply “Good Samaritan”
immunity to a provider that intentionally designs or operates its services in a way that impairs its
ability to identify criminal activity occurring on (or through) its services, or to produce relevant
information to government authorities lawfully seeking to enforce criminal laws. A Good Samaritan is
not someone who buries his or her head in the sand, or, worse, blinds others who want to help.

One important way to confront the grave and worsening problem of illicit and unlawful
material on the internet is to ensure that providers do not design or operate their systems in any
manner that results in an inability to identify or access most (if not all) unlawful content. Such
designs and operation put our society at risk by: (1) severely eroding a company’s ability to detect and
respond to illegal content and activity; (2) preventing or seriously inhibiting the timely identification of
offenders, as well as the identification and rescue of victims; (3) impeding law enforcement’s ability to
investigate and prosecute serious crimes; (4) and depriving victims of the evidence necessary to bring
private civil cases directly against perpetrators.

We propose making clear that, in order to enjoy the broad immunity of Section 230, an
internet platform must respect public safety by ensuring its ability to identify unlawful content or
activity occurring on its services. Further, the provider must maintain the ability to assist government
authorities to obtain content (i.e., evidence) in a comprehensible, readable, and usable format
pursuant to court authorization (or any other lawful basis).

15
 2. Redress for Victims of Terrorism, Child Sex Abuse, and Cyberstalking
Carve-Out for Terrorism Laws

Courts have interpreted Section 230 to apply very broadly, including to block causes of action
under civil anti-terrorism laws. See, e.g., Force v. Facebook, 934 F.3d 53, 68-72 (2d Cir. 2019); Fields
v. Twitter, Inc., 217 F. Supp. 3d 1116, 1118 (N.D Cal. 2016). Given the importance of vigorous criminal
and civil enforcement efforts against terrorism, some have proposed that terrorism laws should be
explicitly carved out from Section 230 immunity. Immunity for knowingly facilitating terrorist content
and activity is far from the core of Section 230, and removing this immunity would not disrupt the
internet and social media ecosystems. Moreover, under a traditional tort standard (which would exist
absent immunity), platforms would only have to take reasonable steps with respect to the presence
of terrorist content; they would not have to achieve perfect success. The Department believes that
exempting narrow categories of egregious conduct, like facilitating terrorism, from Section 230’s broad
immunity strikes the right balance between protecting public safety and national security, on the one
hand, and preserving the core of Section 230 immunity, on the other.

Carve-Out for Child Sex Abuse Laws

Many have similarly proposed to carve out from Section 230 immunity civil enforcement of
federal child sex abuse laws, and civil and criminal enforcement of state child sex abuse laws. (Federal
criminal enforcement of such laws is already carved out.) Immunity for facilitating child sex abuse
and child sexual abuse material (CSAM) is not at the core of Section 230, and removing it would
not disrupt the internet industry, as shown by the statute’s current carve out for human trafficking
content. As with the terrorism carve-out, a tort standard would apply in the absence of immunity, so
platforms that take reasonable steps to address CSAM and child exploitation still would be protected.
The Department supports a tailored carve-out for child sexual exploitation and abuse laws.

Carve-Out for Cyber Stalking

Cyber-enabled stalking is a particularly pernicious and growing threat to victims around the
United States. Cyberstalking includes a course of conduct or series of actions by the perpetrator
that places the victim in reasonable fear of death or serious bodily injury. Prohibited acts include
repeated, unwanted, intrusive, and frightening communications from the perpetrator by phone,
e-mail, or other forms of communications, as well as harassment and threats communicated through
the Internet, such as via social media sites and applications. Given the importance of criminal and civil
enforcement efforts against cyberstalking, the Department would support carving out cyber-stalking
from Section 230 immunity, so that victims can seek civil recourse where platforms fail to exercise due
care to prevent such illicit and harmful behavior.

16
 3. Notice Liability for Federal Criminal Material and Court Judgments
Case-Specific Carve-Out for Actual Knowledge of Federal Criminal Material

As described above, we believe that platforms that purposefully facilitate egregious illicit
activity or material, or are willfully blind to its presence on their services, should be excluded from
immunity generally and, thus, should not be entitled to invoke immunity in any case. Additionally,
the Department believes that platforms that have actual notice of specific criminal material or activity
occurring on their services without taking any action also should not be entitled to immunity in cases
arising from such material or activity.

Traditional tort law recognizes several different forms of intermediary liability for publicizing
the speech of third parties. Publishers, such as newspapers or book publishers, are generally held
strictly liable for defamation they publish as if they were the speaker. Distributors, such as libraries
and newsstands, are held liable only if they knew or should have known the content was unlawful.
And accessories, such as printing presses, are generally not held liable for defamation.

Section 230 provides that internet services shall not be treated as a “publisher or speaker”
of content provided by third parties. Although this language says nothing about distributor liability,
courts have interpreted Section 230 broadly to immunize online platforms from liability arising from
their publication of third-party content even when the providers had actual notice that content was
unlawful and thus could have been held liable at common law as distributors. See, e.g., Zeran v.
America Online Inc., 129 F.3d 327 (4th Cir. 1997) (rejecting argument that Section 230 left “distributor
liability intact” and holding AOL immune even though it had notice of unlawful nature of the postings);
Universal Commc’n. Sys, Inc. v. Lycos, Inc., 478 F.3d 413, 420 (1st Cir. 2007) (“Section 230 immunity
applies even after notice of the potentially unlawful nature of the third-party content.”). The rationale
expressed by courts, and by several participants in the Workshop and Roundtable, is that imposing
notice liability on online platforms would chill speech. Some participants suggested that, if platforms
were subject to notice liability, they would automatically take down any and all content upon notice
without investigation, giving a heckler’s veto to anyone who objects to someone else’s speech.

When it comes to unlawful content related to federal crimes like child exploitation, drug
trafficking, cyber-stalking, or terrorism, however, it is far less clear that we should be concerned about
chilling such activity, and instead should be more concerned about halting such dangerous behavior.
See, e.g., M.A. ex rel. P.K. v. Vill. Voice Media Holdings, LLC, 809 F. Supp. 2d 1041, 1050 (E.D. Mo. 2011)
(holding Backpage.com immune under Section 230 from child sex trafficking claims despite allegations
that the website was “aware of prior cases of minors being sexually trafficked on its website and based
upon the posted ads and photography, no reasonable person could review the postings in the adult
categories and deny prostitution was the object of almost each and every ad”). Several participants at
the Workshop expressed concern that Section 230 did not provide sufficient incentives for platforms
to address such horrendous material, even where they had actual knowledge of its presence and
illegality.

17
 While the Department questions whether the text of Section 230 is properly interpreted as
immunizing platforms in circumstances in which they knowingly distribute unlawful content, it is clear
that a narrower notice liability standard should apply at least in the context of material or activity
of third parties that would constitute a federal crime. If a platform has actual notice of specific
material or activity that is unlawful under federal criminal law, does not remove the material, report
the activity, and preserve related evidence, the platform should not be entitled to immunity for
harm resulting from that specific material. This would be a narrow proposal that would not revoke
immunity for defamation or similar speech torts, thus avoiding the concern over the heckler’s veto. At
the same time, this would provide stronger incentives for platforms to address clearly illegal activity
and material on their services.

Case-Specific Carve-Out To Encourage Compliance with Court Judgments

Some courts have held that Section 230 provides immunity for online platform even when it
fails to take down content that a court has in fact determined to be unlawful. See, e.g., Hassell v. Bird,
420 P.3d 776, 789 (Cal. 2018) (holding that Yelp’s refusal to comply with a court injunction is protected
by Section 230 because its refusal to remove the defamatory material is an “ongoing decision to
publish”). Section 230 should be narrowed so as not to apply in actions where a platform has failed
to take down content or activity, within a reasonable time, after receiving notice that a court in the
United States has adjudicated the content or activity to be unlawful. At the same time, Section 230
should make clear that platforms do have immunity for takedowns consistent with such court orders.

Most online platforms ordinarily comply with court-ordered take-down requests, and such a
clarification of the statute would not impose an undue burden. Victims, however, would benefit from
having a clear path to remove defamation and other unlawful material from platforms, especially in
cases where the underlying poster may not be reachable.

Require Platforms to Provide Users with Mechanisms to Flag Unlawful Content

An online platform is not a Good Samaritan if it sticks its head and the sand and goes out
of its way to avoid receiving notice of criminal content on its service. To ensure a platform has the
ability to be notified of illegal content, interactive computer services should offer an easily accessible
and apparent mechanism for users to give notice to providers so they have an opportunity to review
and remove it where appropriate. Otherwise, there is a risk that a platform might attempt to avoid
receiving actual notice by making it impossible or difficult for users to flag illicit material.

Under current case law interpreting Section 230, courts have held having a mechanism for
users to alert platforms of illegal activity is a voluntary precaution that Section 230 permits but does
not require. See, e.g., Daniel v. Armslist, LLC, 926 N.W. 2d 710, 722 (Wis. 2019). The Department
supports a provision requiring an interactive computer service to have easy and apparent mechanism
for users to flag unlawful content in order to benefit from Section 230 immunity. The mechanism
should be reasonable based on the size and nature of the interactive computer service.

18
 Clarifying Federal Government Civil Enforcement Capabilities
Currently, Section 230 expressly precludes immunity in federal criminal enforcement actions,
see § 230(e)(1), but some have suggested that Section 230 immunity may lie against the federal
government in civil enforcement matters. Recently, the federal government has seen an uptick in
instances in which Section 230 immunity is raised in negotiations over federal civil litigation, or is
invoked as a defense in federal civil enforcement actions. In discussions the Department held with
thought leaders, industry representatives, and legal experts, there was wide agreement that Section
230 should not apply to suits brought by the federal government. Indeed, many were surprised to
hear that this was even an issue the government was facing.

Civil enforcement is an important complement to the federal government’s criminal

prosecutions. Civil actions by the federal government also would not raise the concern of a flood
of private damages litigation over state law claims of defamation that Section 230 in part sought to
address. The Department therefore believes that Section 230 should be amended to make clear that
its immunity does not apply in any case brought by the federal government, whether criminal or civil.

19
 Reform to Promote Competition
A concern that many have raised in the context of Section 230 and more broadly is the
increased size and power of a small handful of online platforms. This is relevant in the Section 230
discussion for those citizens who want safer online spaces, for those whose speech has been banned
or restricted by these platforms, and for upstart businesses trying to compete against these platforms.
Over time, the avenues for sharing information and engaging in discourse with a large number of
individuals have concentrated in the hands of a few key players. Further, the big tech platforms of
today often monetize through targeted advertising and related businesses, rather than charging users.
Thus, their financial incentives in content distribution may not always align with what is best for an
individual user.

Antitrust law prohibits dominant firms from engaging in anticompetitive conduct that harms
competition. In some cases, online platforms have argued that Section 230 creates an immunity from
antitrust claims. See, e.g., Enigma Software Grp. USA, LLC v. Malwarebytes, Inc., 946 F.3d 1040, 1050
(9th Cir. 2019) (rejecting Malwarebytes’s contention that it was immune from liability under Section
230 “regardless of any anticompetitive motives”) (cert pending).

Immunity against antitrust claims, however, was not part of the core objective of Section
230. In an antitrust case, the key question is whether a defendant is engaging in conduct that harms
competition. Such claims are not based on third-party speech, nor do they focus on whether the
platform is a publisher or speaker.

Given this, and the existing market dynamics, it is important to ensure that Section 230 is not
used as a tool to block antitrust claims aimed at promoting and preserving competition. Interpreting
Section 230 based on its text and original purpose does not appear to preclude federal antitrust
claims. However, the Department believes it would be useful to create an explicit legislative carve-out
from Section 230 for claims under the federal antitrust laws. Until then, there is a risk that defendants
will continue to try to use Section 230 creatively to block antitrust actions.

20
 Reforms that Promote Open Discourse and Greater Transparency

1. Replace Vague Language to Address Moderation Beyond Section 230

Currently, Section 230(c)(2) immunizes platforms from liability related to restricting access to,
or the availability of, material that the platforms consider “obscene, lewd, lascivious, filthy, excessively
violent, harassing, or otherwise objectionable.” 47 U.S.C. § 230(c)(2)(A) (emphasis added). Courts
have disagreed over how much discretion platforms have to decide what is “otherwise objectionable.”
Some construe the phrase to confer virtually unlimited discretion on platforms to remove any content
they object to, for whatever reason. See, e.g., PC Drivers Headquarters, LP v. Malwarebytes Inc.,
371 F. Supp. 3d 652, 662 (N.D. Cal. 2019); Langdon v. Google, Inc., 474 F. Supp. 2d 622, 631 (D. Del.
2007). Others counter that such unconstrained discretion would be inconsistent with the policy goals
Congress set forth in Section 230. See, e.g., Enigma Software Grp. USA, LLC v. Malwarebytes, Inc., 946
F.3d 1040, 1049-51 (9th Cir. 2019). Those goals include “preserv[ing] the vibrant and competitive free
market that presently exists for the Internet,” 47 U.S.C. § 230(b)(2), and maintaining the Internet as “a
forum for a true diversity of political discourse,” id. § 230(a)(3).

Unconstrained discretion is particularly concerning in the hands of the biggest platforms,

which today effectively own and operate digital public squares. This is even more salient today where
social distancing requirements have driven more speech and interaction online. The vagueness
of the term “otherwise objectionable” risks giving every platform a free pass for removing any
and all speech that it dislikes, without any potential for recourse by the user. Therefore, to bring
the immunity conferred by (c)(2) more in line with the interests Congress identified in the original
CDA, the Department proposes deleting the vague phrase “otherwise objectionable,” while adding
a new immunity for moderation of material the platform believes, in good faith, violates federal
law or promotes violence or terrorism. By both narrowing and expanding 230(c)(2) in these ways,
the proposals strike a more appropriate balance between promoting an open, vibrant Internet and
preserving platforms’ discretion to restrict obscene and unlawful content.

To be clear, the Department’s proposal would not leave platforms unable to moderate content
on their services. Nor does removal of blanket immunity itself impose liability for content moderation
decisions. Online platforms are often protected by their terms of service when removing content that
violates the platform’s rules, whether or not that content falls into the categories of (c)(2). Therefore,
removing Section 230 immunity from certain content moderation decisions means that platforms
must rely on—and abide by—their terms of service. In our view, incentivizing platforms to be more
transparent and clear in their terms of services, including with respect to content removal decisions,
will ultimately benefit users.

21
 2. Provide Definition of Good Faith
Under subsection (c)(2), platforms must act in “good faith” to receive immunity related to
content-moderation decisions. Several experts have raised the concern, however, that platforms are
ignoring this “good faith” requirement and censoring material in deceptive or pretextual ways.

To address this problem, the Department suggests providing a definition of what constitutes
“good faith.” To restrict access to particular content in “good faith,” a platform should be required to
meet four criteria. First, it must have publicly available terms of service or use that state plainly and
with particularity the criteria the platform will employ in its content-moderation practices. Second,
any restrictions of access must be consistent with those terms of service or use and with any official
representations regarding the platform’s content-moderation policies. Third, any restrictions of access
must be based on an objectively reasonable belief that the content falls within one of the categories
set forth in subsection (c)(2)(A). And fourth, the platform must supply the provider of the content
with a timely notice explaining with particularity the factual basis for the restriction of access, unless
the provider reasonably believes that the content relates to criminal activity or notice would risk
imminent harm to others. These requirements aim to encourage greater transparency in platforms’
content-moderation decisions.

3. Continue to Overrule Stratton Oakmont to Avoid the Moderator’s

Dilemma
Congress enacted Section 230 in part to overrule the Stratton Oakmont decision, in which a
platform was held to be responsible for all content on its service because it chose to moderate some
content on its service. Stratton Oakmont, Inc. v. Prodigy Servs. Co., 1995 WL 323710 (N.Y. Sup. Ct.
1995). Workshop participants expressed concern that reforms to Section 230 might reintroduce the
“Moderator’s Dilemma,” forcing a platform to choose between moderating content (and therefore
exposing itself to liability for all other user-generated content), or not moderating at all (and therefore
hosting whatever content, however repugnant, that users post).

The Department agrees that it is important to avoid the Moderator’s Dilemma and supports
adding a provision to make clear that a platform’s decision to moderate content either under (c)(2)
or consistent with its terms of service does not automatically render it a publisher or speaker for all
other content on its service.

Although a takedown decision either under (c)(2)(A) or consistent with a platform’s terms of
service does not render an online platform a publisher or speaker for all other content, those actions
are treated differently with respect to the statutory immunity. A takedown decision pursuant to (c)
(2)(A) is immune from civil liability under Section 230. A platform’s removal or restriction of content
outside of (c)(2)(A) is not entitled to Section 230 immunity—under either (c)(1) or (c)(2)—even if
consistent with the platform’s terms of service.

22
 PART III: IDEAS UNDER FURTHER CONSIDERATION

1. Distinguish Between Different Types of Internet Services. A number of experts and

industry participants highlighted the distinction between different types of interactive
computer services. Some services, like internet service providers, primarily serve as the
“pipes” for content to flow through and have a more limited ability to detect and report
criminal activity. Section 230’s current definition of “interactive computer services”
may also be interpreted to include collateral services that are not user-facing and do
not themselves host user content. Many of the proposals for reform appear most
concerned with platforms that solicit and curate user-generated content—what some call
“digital curators.” There is an open question on whether and how to vary immunity and
responsibility for the different types of services, especially when technology continues to
change.

2. Distinguish Between First-Party and Third-Party Speech. Several Workshop participants

noted a distinction between when platforms are engaging in their own speech and when
they are hosting third-party speech. Section 230 does not grant immunity to online
platforms from claims arising from information they (rather than a third-party) provide, or
from conduct or design decisions. If a platform is speaking in its own right – or if it edits or
contributes to a third-party post – then it has become an “information content provider”
and unable to seek the protection of Section 230 immunity. See § 230(f)(3). The Section
230 analysis of when a platform becomes an “information content provider,” however,
becomes harder with industry’s use of proprietary algorithms and other technologies,
which blur the line between first and third-party speech.

3. Address Republication Liability. Relatedly, some Workshop participants expressed a

view that Section 230 should not immunize platforms for ratification, republication, or
amplification of unlawful speech. As one participant noted, “freedom of speech is not
freedom of reach.” The question is how to define republication where algorithms and
technology could be seen as “republishing” or “amplifying” almost all speech on the
service. While there may be a way to distinguish where a platform actively promotes
speech on the basis of its substance (e.g., featured story of the day or sponsored content),
this distinction should be carefully considered and defined.

4. Sunsetting May be Appropriate. Given how quickly technology changes, some have
expressed the view that it may be appropriate to sunset Section 230 immunity so that
future legislators must revisit whether technological changes warrant changes to the
scope of Section 230 or whether the immunity is no longer necessary.

24
5. Transparency Reporting Requirements. To foster greater transparency in how platforms
enforce their content moderation policies, some experts have proposed requiring large
platforms to regularly disclose data on the enforcement of their content moderation
policies in order to receive or keep Section 230 immunity. Such disclosure would serve
multiple important aims.

First, it would enable members of the public to identify best practices related to restricting
harmful content. For example, the disclosures could provide data on how aggressively
platforms are enforcing their policies, how they are identifying improper material, and
how long it takes them to remove such material. Scholars, policymakers, and other
platforms could use such information to improve content moderation practices and better
protect against the range of unlawful material that appears online.

Second, disclosure of enforcement data would address concerns that large platforms
discriminate against particular viewpoints in enforcing their content moderation policies.
Currently, platforms have no obligation to disclose data that would enable third-parties
to evaluate whether such bias claims are true. As many experts agreed, access to
robust enforcement data would enable policymakers and the public to evaluate whether
platforms are enforcing content moderation policies even-handedly across different
political viewpoints and communities.

Enforcement data may help to alleviate suspicion of platforms if, as some experts claim,
complaints of bias simply reflect that, given the scale of large platforms, there are many
anecdotal examples that individuals can point to as evidence of bias but that in reality are
not representative of overall content moderation decisions. Alternatively, enforcement
data may help inform consumer choices or policy solutions if they reveal that claims of
bias are well-founded. Either way, public disclosure of robust enforcement data appears
useful to ensuring that the internet remains, in the words of the CDA, “a forum for a true
diversity of political discourse, unique opportunities for cultural development, and myriad
avenues for intellectual activity.” 47 U.S.C. § 230(a)(3).

25