Meta Platforms, Inc. v. District of Columbia

• Court: District of Columbia Court of Appeals
• Decided: September 14, 2023
• Docket: 22-CV-0239
• Status: Published

Opinion

Notice: This opinion is subject to formal revision before publication in the Atlantic and Maryland Reporters. Users are requested to notify the Clerk of the Court of any formal errors so that corrections may be made before the bound volumes go to press.

DISTRICT OF COLUMBIA COURT OF APPEALS

No. 22-CV-0239

META PLATFORMS, INC., APPELLANT,

V.

DISTRICT OF COLUMBIA, APPELLEE.

Appeal from the Superior Court of the District of Columbia (2021-CA-004450-2)

(Hon. Anthony C. Epstein, Trial Judge)

(Argued January 31, 2023 Decided September 14, 2023)

Catherine M.A. Carroll, with whom Ronald C. Machen, George P. Varghese, Ari Holtzblatt, and Joshua S. Lipshutz were on the brief, for appellant.

Ashwin P. Phatak, Principal Deputy Solicitor General, with whom Karl A. Racine, Attorney General for the District of Columbia at the time, Caroline S. Van Zile, Solicitor General, and Stacy L. Anderson, Senior Assistant Attorney General, were on the brief, for appellee.

Before BLACKBURNE-RIGSBY, Chief Judge, DEAHL, Associate Judge, and STEADMAN, Senior Judge.

Opinion for the court by Associate Judge DEAHL.

Concurring opinion by Associate Judge DEAHL at page 42. 2

DEAHL, Associate Judge: The District has subpoenaed Meta Platforms, the

operator of the social media site Facebook, for documents related to Meta’s

enforcement of its COVID-19 misinformation policies. The District is investigating

potential violations of the Consumer Protection Procedures Act, or CPPA, D.C.

Code §§ 28-3901 to -3913, alleging that Meta has misrepresented to the District’s

consumers the degree to which it polices misinformation posted to its platform about

the COVID-19 vaccine. Meta refused to comply with the subpoena, and the Superior

Court issued an order enforcing the subpoena. Meta now appeals that order.

Meta raises two arguments in support of its view that the District’s subpoena

is unenforceable. Its first argument concerns the Stored Communications Act, or

SCA, 18 U.S.C. §§ 2701 to 2711. Meta argues that § 2703 of the SCA requires the

District to procure a warrant in order to compel the disclosure of the documents it

seeks. Its second argument is grounded in the Constitution. Meta argues that the

District’s subpoena infringes on both its and its users’ First Amendment rights to

free speech and free association. Like the trial court, we disagree with Meta as to

both points, and affirm. 3

I.

Superior Court Proceedings

This case arises from an ongoing investigation by Attorney General for the

District of Columbia into Meta’s content moderation practices. Throughout the

COVID-19 pandemic, Meta made various public statements about its efforts to

police the spread of misinformation on its platform. In December 2020, for example,

the company announced that it would be “remov[ing] false claims that COVID-19

vaccines contain microchips, or anything else that isn’t on the official vaccine

ingredient list.” Several months later, Meta unveiled an expansion of this policy,

noting “a particular focus on pages, groups, and accounts that violate these rules.”

By August 2021, Meta reported that these efforts had led to the removal of 20 million

items of content and over 3,000 accounts, pages, and groups for repeat violations.

The District, perceiving a mismatch between these public statements and the

widespread dissemination of vaccine misinformation on Facebook, is investigating

Meta’s potential violations of the CPPA. That statute, which prohibits unfair and

deceptive trade practices, authorizes the District to conduct “investigation[s] to

determine whether to seek relief under” its provisions, including by issuing

subpoenas to “compel production of records, books, papers, contracts, and other 4

documents.” D.C. Code § 28-3910(a). Relying on this authority, the District issued

a subpoena demanding the production of the following:

Documents sufficient to identify all Facebook groups, pages, and accounts that have violated Facebook’s COVID-19 misinformation policy with respect to content concerning vaccines, including the identi[t]y of any individuals or entities associated with the groups, pages, and accounts; the nature of the violation(s); and the consequences imposed by Facebook for the violation, including whether content was removed or banned from these sources.

This demand was eventually narrowed to only those documents related to public

posts, or posts that were so widely accessible as to be functionally public. 1

Meta refused to comply with the subpoena, and so the District brought an

enforcement action in Superior Court. In that litigation, Meta principally argued that

the government may compel the production of electronic communications only by

procuring a warrant, citing to a provision of the SCA, 18 U.S.C. § 2703(a). The trial

court disagreed with that reading of the statute. It instead reasoned that because the

1 It is difficult to say exactly when a post to a nominally private Facebook group or Page has been so broadly disseminated that it is effectively public. The trial court charged Meta and the District with reaching an “agreement on an approach that identifies public posts in a way that protects non-public posts from disclosure and that does not impose an undue burden on Meta.” Neither party challenges that aspect of the trial court’s order, so we do not opine on any theoretical threshold for when a post on the internet becomes functionally public. 5

District is targeting only public posts, the SCA’s “consent exception,” § 2702(b)(3),

permitted Meta to make the disclosures, and Meta was therefore required to comply

with the District’s valid subpoena (more on these provisions in a moment). Meta

also raised a First Amendment challenge to the subpoena, arguing that compelling it

to disclose the targeted documents would chill both its and its users’ First

Amendment rights of free speech and association. The court again disagreed,

concluding that the subpoena did not infringe upon either Meta’s or its users’ First

Amendment rights.

Meta now appeals, pressing the same two arguments that it raised before the

trial court. First, it argues that the SCA precludes the government from compelling

disclosure of the targeted documents via subpoena, as the SCA requires it to instead

procure a warrant. Second, it argues that the subpoena violates its and its users’ First

Amendment rights of free speech and association. We address Meta’s statutory

argument concerning the proper interpretation of the SCA first, and then turn to its

First Amendment challenges.

II.

The proper interpretation of the SCA is a question of law we review de novo.

Facebook, Inc. v. Wint, 199 A.3d 625, 628 (D.C. 2019). 6

A. Background of the Stored Communications Act

Congress passed the SCA in 1986 to fill a perceived hole that technological

advances had poked in the Fourth Amendment’s protections of private

communications and records. For most of our country’s history, people typically

kept their private communications and records in their homes or places of business,

and the government generally needed a warrant supported by probable cause to seize

those materials. See Coolidge v. New Hampshire, 403 U.S. 443, 455 (1971) (the

Fourth Amendment’s warrant requirement is “subject only to a few specifically

established and well delineated exceptions”).

The advent of email and other forms of electronic communications and

storage changed that, and raised serious questions about the Fourth Amendment’s

applications to these new technologies. Electronic communications typically must

be disclosed to third-party service providers, who then transmit messages to their

intended recipients. Those third-party service providers might themselves disclose

the communications to the government, offering a potentially massive end run on

the Fourth Amendment’s protections of private materials. See United States v.

Miller, 425 U.S. 435, 443 (1976) (“[T]he Fourth Amendment does not prohibit the

obtaining of information revealed to a third party and conveyed by [them] to 7

Government authorities.”). But see Carpenter v. United States, 138 S. Ct. 2206,

2217 (2018) (declining to extend Miller to cell-site location information a person

reveals to their wireless carrier). 2

The SCA sought to fill that potential gap by providing “a set of Fourth

Amendment-like privacy protections by statute,” limiting “the ability of [service

providers] to voluntarily disclose information about their customers and subscribers

to the government.” Orin S. Kerr, A User’s Guide to the Stored Communications

Act, and a Legislator’s Guide to Amending It, 72 Geo. Wash. L. Rev. 1208, 1212-

13 (2004). Two of the SCA’s provisions are particularly crucial to this appeal.

First is § 2702, which precludes service providers from disclosing their users’

communications or records, subject to certain exceptions. Section 2702 states that

the provider of an “electronic communication service” may not knowingly divulge

“the contents of a communication while in electronic storage by that service.” 18

2 To be clear, we do not express any view about whether there is in fact any gap in the Fourth Amendment’s protections of electronic communications. Courts are capable of adapting doctrinal rules to fit technological advances, but are often slow to do so. The SCA was simply Congress’s attempt to address the gap it perceived. 8

U.S.C. § 2702(a)(1). 3 The provision then lists nine exceptions to that general

prohibition, including: disclosures “to an addressee or intended recipient of” the

communication, id. § 2702(b)(1); disclosures “with the lawful consent of the

originator or an addressee or intended recipient of” the message, id. § 2702(b)(3);

and disclosures “as otherwise authorized in [§ 2703],” discussed immediately below.

Id. § 2702(b)(2). Section 2702(a) “broadly prohibits providers from disclosing the

contents of covered communications.” Wint, 199 A.3d at 628. But when one or

more of the nine § 2702(b) exceptions apply, we have held that “the SCA is no

obstacle” to compelling disclosure of communications via ordinary legal process,

like subpoenas. Facebook, Inc. v. Pepe, 241 A.3d 248, 253 (D.C. 2020).

Second is § 2703, which confers on government entities alone the power to

compel disclosure of electronic communications and records, even when no

§ 2702(b) exception applies. The SCA grants private parties no similar authority.

Under § 2703, the government may compel via court order a narrow set of non-

3 The SCA’s strictures apply to both providers of “electronic communication services” and “remote computing services.” See 18 U.S.C. § 2510(15) (defining “electronic communication service”), and § 2711(2) (defining “remote computing service”); see also id. § 2702(a)(2) (directed at remote computing services). For our purposes, the differences between these types of services are immaterial, and we refer generally to “service providers.” Cf. Kerr, User’s Guide, 72 Geo. Wash. L. Rev. at 1209 (advocating for “eliminating the[se] confusing categories”). 9

content records, including a subscriber or customer’s name, address, and means of

payment. 18 U.S.C. § 2703(c)(2), (d). As for the contents of electronic

communications, like the text of an email, the statutorily required process for

government-compelled disclosure depends on how long the communication has

been in electronic storage. When it has been in storage for more than 180 days,

§ 2703(b) permits the government to compel its disclosure so long as it provides

prior notice to the user and obtains an administrative subpoena or court order

provided for in § 2703(d). But when the communication has been in storage for 180

days or less, the SCA authorizes the government to compel disclosure “only pursuant

to a warrant.” Id. § 2703(a). 4

B. Wint, Pepe, and the Parties’ Competing Readings of the SCA

We have interpreted these provisions twice before, and both cases are

important here. We first addressed them in Facebook v. Wint, where we held that

4 The District does not invoke the SCA’s more permissive processes for obtaining communications that are “more than one hundred and eighty days” old, 18 U.S.C. § 2703(b), (d), and Meta suggests that is because the 180-day line has effectively been abandoned through practice and case law. Whatever the reason, we proceed here as if the District is seeking materials that have been electronically stored for 180 days or less, even though that would seem not to be true of the vast bulk of materials that the District seeks. 10

§ 2702(a)’s “broad prohibition” on disclosure precludes a service provider from

complying with a criminal defendant’s subpoena seeking protected communications.

199 A.3d at 629-30. None of the statutory exceptions permitting disclosure applied

in that case, and we concluded that “barring an applicable statutory exception, the

SCA prohibits providers from disclosing covered communications,” even when

subpoenaed by a private party. Id. at 629.

We next addressed these provisions in Facebook v. Pepe, where unlike Wint,

statutory exceptions did apply to permit the service provider to disclose the

subpoenaed communications. 241 A.3d 248, 256 (D.C. 2020). Nonetheless,

Facebook opted not to comply with the criminal defendant’s subpoena in that case,

highlighting statutory language providing only that it “may divulge” electronic

communications when such an exception applies, rather than requiring it to do so.

Id. at 257-58 (emphasis added). We disagreed and held that Facebook was required

to comply with the subpoena where nothing in the SCA precluded it from doing so.

Id. at 258. In short, because “the SCA did not authorize Facebook’s refusal to

comply with Mr. Pepe’s subpoena,” Facebook was subject to “disclosure

requirements imposed by other law.” Id. at 258. 11

The present case, like Pepe, involves communications that are exempted from

the SCA’s broad prohibition on disclosure. Meta does not dispute that the SCA

permits it to comply with the District’s subpoena, because the District seeks only

publicly posted messages, which the parties agree fit within the SCA’s consent

exception to overcome the general bar on disclosure. 18 U.S.C. § 2702(b)(3). But

see infra at 42-44 (Deahl, J., concurring) (questioning exception’s application).

The parties offer competing theories about how §§ 2702 and 2703 should be

read together and applied in this case. The District contends that Pepe’s reasoning

applies with full force here, and the same result—that Meta must comply with the

subpoena—follows. Recall that Pepe held that so long as some § 2702(b) exception

to § 2702(a)’s general bar on disclosure applies, private parties can avail themselves

of whatever avenues of compulsory process they have available to them, and service

providers must comply with such valid process. 241 A.3d at 258. The District

maintains that the same is true when a government actor subpoenas documents. In

its view, § 2702(a)’s general prohibition on disclosure does not bar compliance with

its subpoena because the users have consented to the disclosure of their

communications by publicly posting them. See 18 U.S.C. § 2702(b)(3). And as in

Pepe, the District has authority independent of the SCA to “compel production of

records” that it seeks: the CPPA. See D.C. Code § 28-3910. Thus, Meta must 12

comply with the subpoena. The Superior Court adopted essentially this reasoning,

positing that “[n]othing in the text of § 2702(b)(3) limits the consent exception to

disclosure to non-governmental entities.”

Meta counters that Pepe is inapplicable because the SCA applies an entirely

different set of restrictions when it is the government, rather than a private party,

seeking to compel disclosures. Meta contends that Pepe and the § 2702 exceptions

are inapposite in light of § 2703’s directive that “[a] governmental entity may require

the disclosure . . . of the contents of a wire or electronic communication . . . only

pursuant to a warrant.” 18 U.S.C. § 2703(a) (emphasis added). This portion of the

statute, it argues, imposes additional hurdles on government entities seeking

electronic communications beyond those found in § 2702, and the trial court blurred

the distinction between the two provisions when it granted the District’s petition to

enforce its subpoena. If the District or any other government entity wants access to

these records, Meta concludes, it has no option but to comply with § 2703’s warrant

requirement.

We agree with the District’s reading of the statute. The text and structure of

the SCA support the District’s interpretation, as we explain in Part II.C. The Act’s

legislative history also supports that interpretation, as we explain in Part II.D. 13

C. The SCA’s Text and Structure

We begin with the text of the statute, which is “generally the best indication

of the legislative intent.” In re B.B.P., 753 A.2d 1019, 1021 (D.C. 2000). As an

initial matter, the District argues that the SCA does not apply to public posts at all,

but instead applies only “to protect information that the communicator took steps to

keep private.” Ehling v. Monmouth-Ocean Hosp. Serv. Corp., 961 F. Supp. 2d 659,

668 (D.N.J. 2013). It might be right about that. 5 See infra at 44-47 (Deahl, J.,

concurring). But see Facebook, Inc. v. Superior Court (Hunter), 417 P.3d 725, 743-

44 (Cal. 2018) (concluding that the SCA “initially prohibits the disclosure of all

(even public) communications—but that section 2702(b)(3)’s subsequent lawful

5 As the District points out, the SCA’s proscriptions of the unauthorized access and interception of electronic communications do not extend to “electronic communication[s that are] readily accessible to the general public.” 18 U.S.C. § 2511(2)(g)(i); see id. § 2701. Meta counters that we are not concerned here with the unauthorized access or interception of electronic communications, but with disclosures of electronic communications, which are covered by different provisions: §§ 2702 and 2703. Meta’s response is well-taken, but it is not wholly satisfactory. It would be a rather strange statutory regime if the SCA permitted the government (and anybody else) to “intercept” and “access” any and all public posts, while prohibiting the government from compelling disclosure of the exact same material absent a warrant (as Meta’s reading of the statute would dictate). And at a broader level, it would be odd if a statutory regime meant to mimic the Fourth Amendment’s protections swept so much more broadly than the Fourth Amendment itself, and protected communications that had been broadcast to the world. See infra at Part II.D.2. We ultimately do not resolve this broader dispute between the parties, and will assume, without deciding, that the SCA applies to public posts. 14

consent exception allows providers to disclose communications configured by the

user to be public”). We ultimately bypass that question, though, because even

assuming, as the trial court concluded, that the SCA applies to the public posts at

issue here, we agree that the District’s subpoena is enforceable.

As the trial court recognized, our holding in Pepe applies with equal force

when it is the government, rather than a private party, seeking to compel disclosure

of communications that fall within one of the § 2702(b) exceptions. Nothing in the

text of § 2702 suggests a different result. Section 2702’s consent exception to the

general bar on disclosure instructs that a service provider “may divulge the contents

of a communication . . . with the lawful consent of the originator or an addressee or

intended recipient of such communication, or the subscriber in the case of remote

computing service,” drawing no distinctions based on the nature of the recipient. 18

U.S.C. § 2702(b)(3). The parties agree that this exception applies here, and as we

held in Pepe, when the SCA permits the disclosure of electronic communications to

a third party, a service provider must comply with a valid subpoena requiring such

disclosure. 241 A.3d at 258.

Meta counters that Congress enshrined an entirely different set of rules for

government actors seeking to compel disclosure in § 2703. In particular, it 15

highlights the provision’s directive that “[a] governmental entity may require the

disclosure” of electronic communications “only pursuant to a warrant.” 18 U.S.C.

§ 2703(a). Thus, it claims, “a warrant is the sole—exclusive—means by which the

government ‘may require’ disclosure of content.” This textual argument gets off to

a bad start because the plain text of § 2703 does not, in fact, require the government

to obtain a warrant whenever it seeks to compel electronic communications. The

statute requires a warrant only when communications are held in “electronic storage”

by an electronic communication service provider for 180 days or less. 18 U.S.C.

§ 2703(a). Meta ducks this nuance by asserting “the SCA’s original distinction

between communications in storage for more or less than 180 days has largely been

abandoned.” Maybe so, but to the extent that distinction has been abandoned, it was

for constitutional rather than statutory reasons. See, e.g., United States v. Warshak,

631 F.3d 266, 288 (6th Cir. 2010) (“[T]o the extent that the SCA purports to permit

the government to obtain” emails older than 180 days “warrantlessly, the SCA is

unconstitutional.”); H.R. Rep. No. 114-528, at 9 (2016) (citing Warshak and

questioning “constitutional validity” of distinction). So those authorities are no fix

to the initial textual problem with Meta’s statutory interpretation argument.

But the far bigger textual problem with Meta’s interpretation of § 2703 is that

it reads this provision in isolation, whereas “[s]tatutory interpretation is a holistic 16

endeavor.” Grayson v. AT&T Corp., 15 A.3d 219, 238 (D.C. 2011) (en banc)

(citation omitted). When viewed in its broader statutory context, it becomes clear

that § 2703 is an additional grant of authority permitting government actors alone to

compel disclosures even when no exception to § 2702(a)’s broad prohibition on

disclosure applies, not a unique restriction on government actors when a § 2702(b)

exception does apply, as Meta reads it. 6

Start with the fact that government entities are the only actors that the SCA

affirmatively authorizes to compel disclosures of communications covered by

§ 2702(a)’s general prohibition, even when no § 2702(b) exception applies. There

is no similar authorization for private parties to compel disclosures in the face of a

§ 2702(a) bar. So when a § 2702(b) exception applies to lift the bar on disclosure,

it would make no sense if the government’s additional grant of authority could be

weaponized against it, and read to preclude the government from availing itself of

6 We add a reminder that where no § 2702(b) exception applies to exempt the service provider from § 2702(a)’s general prohibitions against disclosure, then compliance with § 2703 really is the government’s exclusive option for compelling disclosures. See Wint, 199 A.3d at 628. And even compliance with the SCA might not be good enough, as the Constitution may provide added protections, for instance, by protecting communications that are held in storage for more than 180 days in a way the SCA’s plain text does not. See Warshak, 631 F.3d at 291 (“[T]o the extent that the SCA purports to permit the government to obtain [emails older than 180] days warrantlessly, the SCA is unconstitutional.”). 17

the same external legal processes that private parties can avail themselves of when

a § 2702(b) exception applies. Through § 2703 Congress provided the government

with an additional tool to compel disclosures that no private party has; it did not erect

an obstacle to disadvantage the government from compelling information that a

private party could obtain. In arguing otherwise, Meta seeks to invert the asymmetry

that the SCA assigns to governmental and private actors’ respective abilities to

compel communications, contrary to the SCA’s overall scheme.

And there are further textual indications that the SCA grants the government

a greater ability to compel protected communications than the average Joe, rather

than less as Meta would have it. For instance, several of the § 2702(b) exceptions

permit the disclosure of otherwise protected communications only to government

entities. Service providers must disclose to the National Center for Missing and

Exploited Children any communications that they become aware of which indicate

a violation of various laws against child pornography. 18 U.S.C. § 2702(b)(6)

(referencing 18 U.S.C. § 2258A’s mandatory reporting requirements). Those

disclosures are required by law, whether or not the government has a warrant.

Providers also may disclose, but only to “a law enforcement agency,” any

communications “inadvertently obtained by the service provider” that “appear to

pertain to the commission of a crime.” Id. § 2702(b)(7). They may disclose only 18

“to a governmental entity” communications that trigger a good faith belief “that an

emergency involving danger of death or serious physical injury” is afoot and requires

“disclosure without delay” to avoid hazardous results. Id. § 2702(b)(8). 7

When §§ 2702 and 2703 are read together, their import is clear: § 2702(a)

broadly precludes service providers from disclosing the contents of their users’

communications, and unless some § 2702(b) exception applies, the government

alone can compel the disclosure of those communications, and can do so only by

complying with § 2703’s strictures. But where a § 2702(b) exception does apply to

remove § 2702(a)’s bar on disclosure, then the government and private parties alike

can avail themselves of the “mandatory disclosure requirements imposed by other

law,” Pepe, 241 A.3d at 258, like the CPPA’s subpoena powers the District invokes

in this case. Section 2703(a) cannot sensibly be read as a bar on the government’s

ability to compel disclosures that private parties could compel, when it is instead an

additional grant of authority to the government that private parties lack. See id. at

7 We have not surveyed the field to examine whether there are some compulsory reporting requirements in state or federal laws that might require disclosure of materials that fall within these §§ 2702(b)(7) or (b)(8) exceptions. If there are, then our reasoning in Pepe would apply there as well, so that service providers are not only permitted to, but must, comply with those requirements (barring some constitutional hurdle to disclosure). 19

257 (noting the “weighty and well-settled presumption against inferring that

Congress silently intended to foreclose or restrict the availability of a core

component of the judicial process such as the subpoena power”).

Meta protests that § 2703 has its own consent exception (applicable only to

non-content records), which it argues would be “wholly superfluous” under our

reading of the statute, as § 2702 also has a consent exception for non-content records.

Compare 18 U.S.C. § 2703(c)(1)(C) with id. § 2702(c)(2). That’s wrong. The

§ 2702(c)(2) consent exception removes a bar or disability on the service provider

that would otherwise preclude them from disclosing such records, whereas

§ 2703(c)(1)(C) is an affirmative authorization permitting the government to compel

disclosure when the user consents to it. Put another way, when a § 2702 exception

applies, the government—like any other party—can compel disclosure only if they

can point to some authority that allows them to do so; here the District points to its

authority to subpoena records and compel disclosures in furtherance of a CPPA

investigation. See D.C. Code § 28-3910. Whereas when § 2703(c)(1)(C) authority

applies, the government need not point to some external source of authority that

permits it to compel disclosure of non-content records—the provision itself provides

that. There is no superfluity. 20

Meta next counters that our reading of the SCA’s text would make us “the

sole outlier” among courts to have considered this issue. That is a rhetorical sleight

of hand. Meta points to just two decisions from trial courts that it suggests support

its view, and only one of them even arguably does. In truth, Meta’s position here is

so novel that there are simply not any appellate court decisions addressing it, and the

trial court decisions that Meta cites give us no pause.

Meta first points to FTC v. Netscape Communications Corp., 196 F.R.D. 559

(N.D. Cal. 2000). That case is inapposite because there was no suggestion in it that

a § 2702(b) exception applied to the communication sought to be compelled via

agency subpoena. Netscape thus stands for the unremarkable position that where no

§ 2702(b) exception applies, the government’s sole recourse for compelling

disclosure is to comply with § 2703’s strictures. We agree with that—we held

likewise in Wint, 199 A.3d at 629—but it is not the issue before us.

Meta’s other authority is closer to the mark, but unpersuasive. See People v.

Harris, 949 N.Y.S.2d 590 (N.Y. Crim. Ct. 2012). Harris involved a subpoena issued

by a District Attorney’s Office seeking tweets publicly posted from a Twitter

account, allegedly operated by a criminal defendant, over the course of more than

100 days. Id. at 591. The trial court enforced that subpoena as to all but a single 21

day’s tweets, because only that day’s tweets were “less than 180 days old,” and

therefore the court concluded they could be compelled only by a search warrant. Id.

at 596, 598. The court did not address whether the tweets fell within any § 2702(b)

exception to the SCA’s general bar on disclosure, and there is no indication that any

party raised that point. It did not grapple with the statutory structure of the SCA, as

discussed above. And unsurprisingly the parties and the court alike were more

focused on (1) the 100-plus days of tweets that fell outside the 180-day window and

therefore did not require a warrant than they were with the single day of tweets that

fell within it, and (2) the attendant question of whether Twitter users’ themselves

had standing to quash the subpoena served on Twitter. Id. at 593. As to the

remaining single day’s tweets, the court offered nothing resembling persuasive

statutory analysis, stating only that “the government must obtain a search warrant

for the December 31, 2011 tweets.” Id. at 596. That conclusion is some support for

Meta’s view here, but it is anemic, and it gives us no cause to reconsider our own

statutory analysis.

D. The SCA’s Legislative History

The legislative history supports our reading of the SCA as well. As previously

explained, the SCA is roughly meant to extend Fourth Amendment protections to 22

electronic communications and the like. It seeks to neutralize the incident of

technology that things like emails are typically disclosed to third-party service

providers—thereby calling the Fourth Amendment’s protections into doubt—for

purposes of transmission. There are two features of Meta’s proposed interpretation

of the SCA that do not square with this history: (1) it would extend Fourth

Amendment-like protections to public disclosures, which would ordinarily receive

slim-to-no Fourth Amendment protections; and (2) the protections it extends would

not actually belong to the individual users themselves, but instead would belong to

the service providers. We elaborate below on why neither feature aligns with the

legislative history, then we respond to Meta’s counterpoints to it, but before all of

that, we detail the legislative history itself.

1. The Legislative History in Broad Strokes

The SCA was enacted as part of the Electronic Communications Privacy Act

(ECPA) of 1986, Pub. L. No. 99-508, which predates the World Wide Web by

several years. As one might expect, applying the SCA to modern technology is often

like cramming a square peg into a round hole. See generally Orin Kerr, The Next

Generation Communications Privacy Act, 162 U. Pa. L. Rev. 373, 378, 390-410 23

(2014) (detailing a variety of reasons “why the [ECPA] is based on outdated

assumptions”).

As the Senate Report accompanying the legislation explained: “When the

Framers of the Constitution acted to guard against the arbitrary use of Government

power to maintain surveillance over citizens, there were limited methods of intrusion

into the ‘houses, papers, and effects’ protected by the fourth amendment.” S. Rep.

No. 99-541, at 1-2 (1986). Because of technological developments, however,

Congress believed that the Constitution’s protections had become “hopelessly out of

date.” Id. at 2. Unlike one’s physical property, electronic records and

communications are frequently in the possession and control of third-party service

providers, which arguably renders them “subject to no constitutional privacy

protections.” Id. at 3 (citing Miller, 425 U.S. 435). “Thus, the information may be

open to possible wrongful use and public disclosure by law enforcement authorities

as well as unauthorized private parties.” Id. The SCA sought to fill this perceived

gap in the Fourth Amendment’s protections:

[T]he law must advance with the technology to ensure the continued vitality of the fourth amendment. Privacy cannot be left to depend solely on physical protection, or it will gradually erode as technology advances. Congress must act to protect the privacy of our citizens. If we do not, we will promote the gradual erosion of this precious right. 24

Id. at 5; see also H.R. Rep. 99-647, at 19 (“Additional legal protection is necessary

to ensure the continued vitality of the Fourth Amendment.”).

2. Meta Would Expand the SCA Far Beyond the Fourth Amendment

This Congressional intent—to eliminate an instance of legal arbitrage by

applying the Fourth Amendment’s protections to a new technology via statute—

comports with our reading of the SCA’s disclosure provisions. Communications

blasted in public fora, for all to see or hear, generally are not protected by the Fourth

Amendment, putting the nicety of third-party electronic transmitters of

communications aside. “What a person knowingly exposes to the public, even in his

own home or office, is not a subject of Fourth Amendment protection.” Katz v.

United States, 389 U.S. 347, 351 (1967); accord Biles v. United States, 101 A.3d

1012, 1024 (D.C. 2014). So it would make little sense to extend the SCA’s

protections to such communications.

Meta’s interpretation would do just that, despite the fact that publicly

broadcast communications have no shelter in the Fourth Amendment itself. That

would expand the SCA far beyond the Fourth Amendment protections that Congress

sought to mimic. To the contrary, the House and Senate Reports affirmatively

indicate that Congress did not intend for the SCA’s protections to cover content that 25

the user took no steps to keep private. For example, both reports include extended

discussions of electronic bulletin board systems (BBS)—“early analogues to the

social media platforms at issue here.” Hunter, 417 P.3d at 739. As one article

describes this archaic technology, dialing into a BBS was akin to “visit[ing] the

private residence of a fellow computer fan electronically. BBS hosts had converted

a PC . . . into a digital playground for strangers’ amusement.” Benj Edwards, The

Lost Civilization of Dial-Up Bulletin Board Systems, The Atlantic (Nov. 4, 2016).

These early (now anachronistic) digital meeting spaces could be configured as either

“public or semi-public in nature, depending on the degree of privacy sought by

users.” S. Rep. 99-541, at 9. Only the latter, non-publicly accessible BBSs were

intended to fall within the SCA’s protections. As the Senate Report puts it, the

SCA’s protections do not apply where a BBS “does not require any special access

code or warning to indicate that the information is private. To access a

communication in such a public system is not a violation of the Act, since the general

public has been ‘authorized’ to do so by the facility provider.” Id. at 36; see also

Snow v. DirectTV, Inc., 450 F.3d 1314, 1321 (11th Cir. 2006) (“[T]he requirement

that the electronic communication not be readily accessible by the general public is

material and essential to” the SCA’s scope of protections). 26

Meta counters that this discussion of BBSs relates only to the SCA’s

provisions prohibiting the unauthorized access and interception of electronic

communications, or what is effectively the SCA’s anti-hacking provision, 18 U.S.C.

§ 2701. That is not quite right. BBSs feature heavily in the House Report’s

discussion of § 2702, one of the two disclosure provisions we are concerned with

here. Specifically, after noting that a user can waive the SCA’s protections by

consenting to their communications’ disclosure, the report states that “a subscriber

who places a communication on a computer ‘electronic bulletin board,’ with a

reasonable basis for knowing that such communications are freely made available to

the public, should be considered to have given consent to the disclosure or use of the

communication.” H.R. Rep. 99-647, at 66; see also 18 U.S.C. § 2702(b)(3). In other

words, Congress clearly contemplated that publicly broadcast communications

would not be protected under § 2702’s broad prohibition on disclosure. While Meta

is correct that BBSs were not specifically discussed in relation to § 2703, what is

missing from the legislative history is any indication whatsoever that Congress

intended to preclude the government from obtaining, via subpoena or other

compulsory process, materials that were not protected under § 2702 in the first place. 27

3. Meta Would Leave the SCA’s Protections to Service Providers’ Discretion

There is another feature of Meta’s statutory interpretation that is at odds with

the SCA’s legislative history. That history evinces Congress’s intent to confer upon

individual users of electronic services Fourth Amendment-like protections. But

recall that Meta’s view is not that its users have any right to prevent it from

complying with the subpoena in this case—because a § 2702(b) exception applies,

Meta acknowledges that it is free to comply with the subpoena. Meta’s view is

instead that it alone decides whether it will comply with, or defy, the subpoena,

entirely at its own discretion. Under that view, the SCA in fact confers no

protections to Meta’s users when a § 2702(b) exception applies, save for those that

their service provider’s good graces—and maybe the terms of service—afford them.

There is simply nothing in the SCA’s legislative history that suggests Congress

meant to enshrine such service-provider-centric protections. 8 To the contrary, by

8 We do not mean to overstate the point, because users could enter into private agreements with their service providers—like agreeing to a social media site’s terms of service—that preclude disclosure of their private information. It might be a sensible enough regime to leave users’ privacy interests to such agreements, and we do not opine on that. Cf. Orin Kerr, Terms of Service and Fourth Amendment Rights (Jan. 29, 2023), U. Penn. L. Rev. (forthcoming), available https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4342122; https://perma.cc/7MDY-RJ7B (arguing that terms of service do not generally alter Fourth Amendment rights). We conclude only that there is no indication in the 28

seeking to mirror the Fourth Amendment’s protections, the legislative history

evinces Congress’s intent to protect individual users from the discretionary

disclosure choices that their service providers might otherwise make.

It is true that when a § 2702(b) exception applies to permit the disclosure of

otherwise protected communications, the statutory text itself says only that the

provider “may divulge the contents of a communication,” which would generally

connote some degree of discretion, consistent with Meta’s view. But we have

already explained in Pepe why that generally permissive statutory language is a bit

of a mirage: “[T]he subdivisions in § 2702 where ‘may’ appears are framed not as

a grant of discretionary power but as a special exception to a general prohibition. In

such a context all ‘may’ means is that the actor is excused from a duty, liability, or

disability,” it does not “suggest unlimited discretion.” 241 A.3d at 258 (cleaned up)

(first quoting Hunter, 417 P.3d at 751, then quoting Zadvydas v. Davis, 533 U.S.

678, 697 (2001)). And once the § 2702(a)’s general prohibition on disclosure is

lifted, via a § 2702(b) exception, the provider’s discretion is subject to “disclosure

requirements imposed by other law.” Id.

legislative history that this is the regime that Congress had in mind when passing the SCA. 29

4. Meta’s Counterpoints Are Unpersuasive

Meta offers two counterpoints from the legislative history, but neither is

persuasive. First, it notes that when discussing § 2703, the Senate Report states that

“[a] government entity can only gain access to the contents of such an electronic

communication pursuant to a warrant.” S. Rep. 99-541, at 38 (emphasis added).

Similarly, the House Report describes § 2703 as providing “the procedures the

government must use before it can obtain access to the contents of any electronic

communication held by a provider of a remote computing service.” H.R. Rep. 99-

647, at 67 (emphasis added). But while Meta argues that these statements “could

not be more clear,” they in fact do not contemplate communications that are

unprotected by § 2702 in the first place (owing to the applicability of § 2702(b)

exception). Quite the opposite. The premise underlying these discussions was “that

the contents of [the] message in storage were protected by the Fourth Amendment,”

H.R. Rep. 99-647 at 68, and Congress was of the correct understanding that publicly

disclosed communications received no such Fourth Amendment protections, as the

history detailed above makes clear. This history thus supports our view, that where

communications are not protected by § 2702’s anti-disclosure provisions in the first

place, § 2703’s warrant requirement does not apply. 30

Next, Meta argues that applying § 2702’s exceptions to cases involving

government subpoenas ignores Congress’s stated intent to “guard against the

arbitrary use of Government power to maintain surveillance over citizens.” S. Rep.

99-541, at 1. That was certainly Congress’s intent, but our reading of the statute

comports with rather than ignoring it. As discussed, it was well established at the

time of the SCA’s enactment that the Fourth Amendment generally does not protect

the privacy of information that an individual has broadcast to the public. Katz, 389

U.S. at 351. That describes the electronic communications at issue in this case.

There is nothing arbitrary about giving the government the ability to compel the

disclosure of such publicly broadcast communications in much the same way that a

private citizen might do. Pepe, 241 A.3d at 258. It would seem far more arbitrary

to preclude the government from compelling disclosures that any private citizen

might extract. Meta points us to nothing in the legislative history to suggest that

Congress meant to put the government in an inferior position, vis-à-vis private

parties, to compel such disclosures. And we detect no hint of that notion in the

legislative history ourselves.

* * * 31

In summary, the SCA does not authorize a service provider’s refusal to

comply with valid legal process seeking material that a § 2702(b) exception permits

it to divulge. Pepe, 241 A.3d at 258. Because the SCA permits Meta’s compliance

with the District’s valid subpoena, it must comply, as there “is no reason to think the

SCA . . . preempts laws that require disclosures the SCA expressly permits.” Id.

Section 2703 cannot sensibly be read to uniquely inhibit the government’s ability to

compel disclosures that any private party could compel, as Meta contends. The

SCA’s text, structure, and legislative history point to the opposite conclusion: § 2703

was a unique grant of authority to the government—one granted to no private

party—to override § 2702(a)’s broad prohibition in certain circumstances, not a

unique disability on the government when a § 2702(b) exception already applies to

lift that broad prohibition. Meta’s contrary view would stand that statutory scheme

on its head.

III.

We now turn to Meta’s argument that the District’s subpoena impermissibly

intrudes on both its and its users’ First Amendment rights of free speech and

association. 32

We begin by laying some legal groundwork. A court will ordinarily enforce

an investigative subpoena so long as it meets the three-prong test announced in

United States v. Morton Salt Co., 338 U.S. 632, 652-53 (1950). Under that test,

“[w]e consider only whether [1] ‘the inquiry is within the authority of the agency,

[2] the demand is not too indefinite and [3] the information sought is reasonably

relevant.’” Resol. Tr. Corp. v. Thornton, 41 F.3d 1539, 1544 (D.C. Cir. 1994)

(quoting Morton Salt, 338 U.S. at 652). There is no dispute, and we agree, that this

test is satisfied here: (1) the District, through its Office of the Attorney General, is

charged with enforcing the CPPA and may “subpoena witnesses” and “compel

production of records” under its investigative authority to do so, D.C. Code

§ 28-3910(a); (2) its demands are not indefinite; and (3) the information it seeks to

compel is reasonably relevant to its investigation.

But Meta argues that Morton Salt does not apply here, because where

compelled disclosures seriously implicate First Amendment interests, government

subpoenas may face more exacting judicial scrutiny. The seminal case for this

proposition is NAACP v. Alabama ex rel. Patterson, 357 U.S. 449 (1958), which

involved an attempt by Alabama’s attorney general to compel the disclosure of the

NAACP’s membership lists. Id. at 452. The NAACP refused to comply, and it was

held in civil contempt and fined $100,000. Id. at 453-54. The Supreme Court of the 33

United States reversed, reasoning that Alabama’s investigation into the NAACP

“entail[ed] the likelihood of a substantial restraint upon the exercise by petitioner’s

members of their right to freedom of association,” and that Alabama had failed to

demonstrate an interest “which is sufficient to justify the deterrent effect.” Id. at

462-63. The Court more recently described this “exacting scrutiny” standard as

requiring “a substantial relation between the disclosure requirement and a

sufficiently important governmental interest.” Ams. for Prosperity Found. v. Bonta,

141 S. Ct. 2373, 2383 (2021) (“AFPF”) (quoting Doe v. Reed, 561 U.S. 186, 196

(2010)). 9

But that more recent decision in AFPF did not suggest that all government

subpoenas are doomed under the exacting scrutiny standard, regardless of the nature

of the information sought. Rather, the Supreme Court indicated that compelled

disclosures need only satisfy this standard when “First Amendment activity is

chilled—even if indirectly.” Id. at 2384. When it is not, the far more deferential

9 Meta suggests in a single footnoted sentence that because the District’s subpoena seeks a “content-based disclosure,” it should be subject to strict, rather than merely exacting, judicial scrutiny. But a majority of the Supreme Court recently rejected that more demanding standard in AFPF. See 141 S. Ct. at 2383 (Roberts, C.J.) (three justice plurality employing exacting scrutiny); see id. at 2396 (Sotomayor, J., dissenting) (agreeing with the plurality that exacting scrutiny applies). Meta makes no actual argument as to why that majority view should not control here, so we apply it. 34

Morton Salt standard continues to govern our review. A party such as Meta claiming

a First Amendment privilege bears the burden of “demonstrat[ing] that enforcement

of the discovery requests will result in . . . consequences which objectively suggest

an impact on, or ‘chilling’ of, the members’ associational rights.” Perry v.

Schwarzenegger, 591 F.3d 1147, 1160 (9th Cir. 2010) (citation omitted). Only after

this prima facie showing do we consider if a subpoena satisfies exacting scrutiny.

Id. at 1161; accord In re Motor Fuel Temperature Sales Pracs. Litig., 641 F.3d 470,

488 (10th Cir. 2011) (“[T]he party claiming a privilege always bears the initial

burden of establishing the factual predicate for the privilege.”).

For the reasons that follow, we conclude that Meta has not shown that the

District’s subpoena, which seeks information related to publicly accessible content

generated by its users, will result in chilling Meta’s free speech or associational

rights. As to Meta’s users, we assume the exacting scrutiny standard applies, but

conclude that the District has demonstrated that its subpoena is “narrowly tailored

to the government’s asserted interest.” AFPF, 141 S. Ct. at 2383. We therefore hold

that enforcing the District’s subpoena does not violate the First Amendment. 35

A. Meta’s Own First Amendment Rights

We begin with Meta’s claim that the District’s subpoena impermissibly

intrudes upon its own First Amendment rights by “prob[ing] and penaliz[ing]” its

ability to exercise editorial control over the content that is disseminated through its

platform. The trial court disagreed, concluding that even if the First Amendment

protects the ability of a private social media company to make unfettered content

moderation decisions, 10 enforcing the District’s subpoena would not chill Meta from

engaging in that activity, so that exacting scrutiny is unwarranted. We agree.

At its core, Meta’s argument boils down to two assertions: that the District’s

investigation (1) is really just an attempt to “pressure Meta into changing how it

exercises [its] protected editorial control over its platform”; and (2) that government

scrutiny of its practices more generally will lead to a chilling of the company’s

speech.

10 It is far from clear that it does. Federal courts are sharply divided—in multiple senses of the phrase—on the point. Compare NetChoice, LLC v. Attorney General, 34 F.4th 1196, 1210 (11th Cir. 2022) (holding that the First Amendment protects a social media platform’s right to moderate user-generated content as it sees fit), with NetChoice, LLC v. Paxton, 49 F.4th 439, 445 (5th Cir. 2022) (holding that it does not, “reject[ing] the idea that corporations have a freewheeling First Amendment right to censor what people say”). The District does not press the issue, however, so we assume without deciding that this First Amendment right does exist. 36

On the first point, we disagree with Meta’s characterization of the District’s

investigation. As the subpoena itself states, the District is investigating only whether

Meta’s “representations regarding efforts to prevent and remove vaccine

misinformation from the Facebook platform” violate the District’s consumer

protection statute, the CPPA. There is no suggestion that the District is investigating

whether Meta’s moderation policies or efforts to police them were unlawful or

insufficient in themselves (except to the extent that they belie Meta’s

representations). The District has disclaimed any interest in regulating Meta’s

editorial judgment when it comes to its content moderation, and Meta’s reply brief

expressly denies accusing the District of acting in bad faith. This was a prudent

concession. While it is certainly possible for an otherwise valid government

investigation to be launched on pretexutal grounds, Meta points to no evidence that

this is the case here. See Dep’t of Com. v. New York, 139 S. Ct. 2551, 2573–74

(2019) (emphasizing that a “strong showing of bad faith or improper behavior” is

required before inquiring whether an agency is acting pretextually (citation

omitted)).

As to Meta’s argument that the District’s subpoena (even if issued as part of

a legitimate investigation) nonetheless chills its speech, we again disagree. To

reiterate, the only speech that is being targeted by the District’s investigation are 37

Meta’s public statements regarding the company’s content moderation practices,

which the District alleges were deceptive and in violation of the CPPA. If those

allegations are true, then an enforcement action under the CPPA would pose no

constitutional problem at all, as the First Amendment “does not prohibit the State

from insuring that the stream of commercial information flow cleanly as well as

freely.” Va. State Bd. of Pharmacy v. Va. Citizens Consumer Council, Inc., 425 U.S.

748, 772 (1976). In other words, even if content moderation is itself protected

speech, fraudulent misrepresentations regarding a company’s moderation practices

is not.

Meta tries to take this argument a step further, claiming an investigation into

its statements about its content moderation practices might indirectly chill those

practices themselves. “[J]ust as a subpoena demanding notes from an editorial board

meeting would risk chilling a newspaper’s editorial rights,” Meta argues, so too does

the subpoena here threaten its “exercise of editorial control.” The problem with that

analogy is that Meta not only made its content moderation policies publicly

available, it then widely touted the actions that were supposedly taken pursuant to

those policies; indeed, those public statements were the basis for the District’s

investigation. To piggyback on the editorial board analogy, if the newspaper itself

had published an account of its editorial policies and decisions, and it turned out to 38

be potentially fraudulent in some way, it would not chill the newspaper’s exercise

of editorial control to investigate whether the newspaper’s public statements on that

topic were false. Meta offers no theory for how a subpoena targeting documents that

tangentially relate to this entirely public information risks any chilling of its speech,

and we likewise discern none.

B. Meta’s Users’ First Amendment Rights

Meta also argues that enforcing the District’s subpoena would chill the First

Amendment rights of its users. In essence, its theory is that forcing Meta to identify

the users whose posts were removed under the company’s COVID-19

misinformation policy “associate[s]” those users with “speech that [the District]

views as undesirable.” That association, Meta argues, risks deterring these users

from engaging in future online discussions of controversial topics. See Talley v.

California, 362 U.S. 60, 65 (1960) (“[I]dentification and fear of reprisal might deter

perfectly peaceful discussions of public matters of importance.”). We seriously

doubt that. The District seeks disclosures related to public posts, and the users who

made those posts have already openly associated themselves with their espoused

views by publicly posting them to Facebook. While we doubt exacting scrutiny 39

should apply here, we will assume that it does for the sake of argument, and conclude

that the District’s subpoena nonetheless passes constitutional muster.

Recall that exacting scrutiny examines the fit between the importance of the

government’s interest and the means used to realize that interest. “To withstand this

scrutiny, the strength of the governmental interest must reflect the seriousness of the

actual burden on First Amendment rights.” AFPF, 141 S. Ct. at 2383 (quoting Doe,

561 U.S. at 196). More concretely, there must be “a substantial relation between the

disclosure requirement and a sufficiently important governmental interest, and . . .

the disclosure requirement [must] be narrowly tailored to the interest it promotes.”

Id. at 2385 (citations omitted).

The District’s subpoena satisfies both of these requirements. The CPPA

“establishes an enforceable right to truthful information from merchants about

consumer goods and services that are or would be purchased, leased, or received in

the District of Columbia.” D.C. Code § 28-3901(c). And its list of prohibited trade

practices includes instances where a company “misrepresent[s] . . . a material fact

which has a tendency to mislead.” Id. § 28-3904(e). While the merits of the

District’s investigation are not presently before us, it seems plausible at first blush

that false or misleading statements regarding a social media company’s attempts to 40

control the spread of COVID-19 misinformation might run afoul of this statute. As

a result, we are satisfied that any First Amendment impact resulting from the

District’s investigation is in service of a sufficiently important government interest.

Indeed, as AFPF itself held, “[i]t goes without saying that there is a ‘substantial

governmental interest[] in protecting the public from fraud.’” 141 S. Ct. at 2386

(quoting Vill. of Schaumburg v. Citizens for a Better Env’t, 444 U.S. 620, 636

(1980)).

Meta acknowledges that the District has a “legitimate interest in consumer

protection in general.” It nonetheless argues that such an interest is not implicated

here, where Meta’s public statements about content moderation were mere “puffery”

and therefore non-actionable under the CPPA. But commercial puffery is non-

actionable because it consists of statements whose “truth or falsity . . . cannot be

precisely determined,” such as a sign in a storefront window promising “Satisfaction

Guaranteed.” Pearson v. Chung, 961 A.2d 1067, 1076 (D.C. 2008) (quoting

Tietsworth v. Harley-Davidson, Inc., 677 N.W.2d 233, 245 (Wis. 2004)). This sort

of general assertion, incapable of measurement, is unlikely to lead reasonable

consumers astray and therefore cannot be the basis for a CPPA violation. Id. But

that does not describe Meta’s public statements about its COVID-19 misinformation

policy. Meta claimed that it removed 20 million items of content and over 3,000 41

user accounts as a result of enforcing that policy. These are not the “[l]ofty but

vague” statements that can be chalked up to puffery. See Prager Univ. v. Google

LLC, 951 F.3d 991, 1000 (9th Cir. 2020). They are instead quite detailed,

quantifiable, and capable of verification.

As to the fit between that government’s interest and the scope of the District’s

investigation, we likewise conclude that the subpoena—now that it has been limited

to documents relating to publicly accessible posts—is sufficiently tailored. Though

Meta claims that the District’s subpoena could have pursued “less intrusive

alternatives,” such as aggregated or anonymized data, 11 some of the statements that

are the target of the District’s investigation concern the company’s actions regarding

repeat offenders and individuals publicly identified as major purveyors of COVID-

19 misinformation. The investigation focuses not on the users spreading

misinformation or the specific content of their public posts, but on Meta’s statements

11 The trial court’s order does not require Meta to “unmask” any anonymous users, as it requires Meta to produce “only the identities that these users themselves employed in public posts.” Meta counters that even that order might “chill protected speech by disclosing users to the government who identified themselves only to ‘private groups.’” But recall that these groups are only nominally private, and the trial court’s order targets information regarding posts that were spread so widely as to be functionally public. It is hard to see how a user who broadcasts their posts so widely would be chilled by disclosure here (when any recipient of the broadcast could have disclosed the posts to the government themselves). 42

about its regulation of that misinformation. The Superior Court found there is not a

“less intrusive means” for the District to carry out this investigation than the

subpoena at issue, and we likewise see none. Accordingly, because the subpoena is

appropriately tailored to serve the government’s interest, and that interest is

sufficiently important, it satisfies exacting scrutiny.

IV.

For the foregoing reasons, the judgment of the Superior Court is affirmed.

So ordered.

DEAHL, Associate Judge, concurring: I am in full agreement with the court’s

opinion and write separately to address two issues it rightly bypasses. First is

whether § 2702(b)(3)’s consent exception actually applies on the facts of this case.

Second, taking a step back, is whether the SCA’s protections apply to publicly

posted messages at all. I think both of those questions should be answered in the

negative, which would lead to the same result the court reaches: Meta must comply

with the District’s subpoena. 43

First, on the question of consent, I adhere to a general rule of thumb when

trying to figure out if somebody consents to something: You ask them. Here, the

users whose posts are targeted by the District’s subpoena have not been asked

whether they consent to disclosure to the government, and so I find it artificial to say

they have consented to such disclosure. The trial court reasoned to the contrary, that

“when a user posts content on Facebook that is generally accessible to the public,

the user implicitly consents to disclosure.” While that might be a fair inference if

the posts remained public, the posts at issue here have all been removed, so I see no

reason to conclude that any consent to disclosure endures. People are generally free

to withdraw consent and might do so by, for instance, removing or restricting access

to a once-public post. See Ford v. United States, 245 A.3d 977, 984-85 (D.C. 2021)

(recognizing ability to withdraw or revoke consent to a search). The fact that the

posts at issue here are no longer public would preclude me from inferring any present

consent to disclosure. Nonetheless, Meta does not challenge this aspect of the trial

court’s ruling, and so I agree with the opinion for the court that this point has been

conceded. Supra at 11-13.

To be sure, a person who publicly posts something opens themselves up to the

risk—really, the high likelihood when it comes to popular social media sites—that

some third party will save their post for posterity and render any attempt to delete it 44

from public viewing futile. But that is just to say that third parties generally may do

what they will with publicly disclosed communications, which is quite different from

saying the user consents to whatever they do. And a service provider is not free to

do what they want with the communication if the SCA’s protections apply to it (an

important caveat discussed next): they are constrained by the statute, and where the

statute requires the user’s consent to disclosure, I do not think that eternal consent

can be fairly inferred from the fact that a person once publicly posted something.

Second, I agree with the District that the SCA does not apply to public posts

in the first place, 1 so my above concern with the consent exception’s application is

an entirely academic point here. The SCA was enacted to “protect electronic

communications that are configured to be private.” Konop v. Hawaiian Airlines,

Inc., 302 F.3d 868, 875 (9th Cir. 2002). When read as a whole, and in light of the

legislative history discussed extensively in the court’s opinion, the SCA’s apparent

1 The court correctly does not reach this weighty issue because it is obviated by Meta’s concession that the consent exception applies. I note that, in advancing its view that the SCA does not protect public posts, the District places too much weight on 18 U.S.C. § 2511(2)(g)(i), which it suggests is an unambiguous standalone textual basis for concluding that the SCA does not apply to public posts. It is not. Meta correctly counters that this provision concerns the intercept or access of electronic communications, not their disclosure. Still, the provision is some evidence that the SCA was not meant to reach public posts, and the Act’s overall structure and legislative history provide much more evidence for that conclusion. 45

“purpose is to protect information that the communicator took steps to keep private.”

Ehling v. Monmouth-Ocean Hosp. Serv. Corp., 961 F. Supp. 2d 659, 668 (D.N.J.

2013); Snow v. DirecTV, Inc., 450 F.3d 1314, 1321 (11th Cir. 2006) (“[T]he

requirement that the electronic communication not be readily accessible by the

general public is material and essential to” the scope of the SCA’s protections.).

When a person publicly posts a message for the world to see, it falls outside

of the SCA’s protections altogether. In that case, the service provider is best seen as

providing a public platform for a user to broadcast a message, rather than acting as

an “electronic communication service,” a phrase the statute seems to use to refer to

a third-party transmitter of otherwise private communications.

This reading of the SCA makes sense. The SCA was meant to effectively

neutralize the undesired but necessary disclosure of private communications to third-

party service providers; it is not as if the user wants to share their communications

with service providers, so much as they are necessary conduits for relaying messages

to their intended recipients. The SCA steps into that relationship to dictate that the

disclosure to a third-party service provider merely for the purposes of transmitting

the message is a non-event, and should not affect the user’s privacy interests in their

communications that might otherwise be deemed private. But when a user blasts a 46

message for the world to see, the service provider does not act merely as a necessary

transmitter of that communication, but can itself be seen as a recipient of it (just like

everybody else). The third-party transmittal problem that is the SCA’s raison d’être

no longer exists. In that situation there is no Fourth Amendment gap for the SCA to

fill, so it makes little sense to extend the SCA’s protections to it.

My view admittedly faces a textual hurdle, which is that nothing in the

statutory definitions of “electronic communication” or “electronic communication

service” expressly says that the communication at issue must be a private one. See

18 U.S.C. § 2510(12), (15). Those capacious definitions in fact suggest otherwise.

But that is unsurprising given that the SCA was passed in 1986 and there simply

were no platforms for publicly posting electronic messages for the world to see, at

least not on anywhere near the scale of what is available today. The issue was not

on the forefront of legislators’ minds. The closest analogues to social media

platforms at the time were fairly obscure electronic bulletin board systems, which

were analogous in only the barest of ways, and the limited legislative history on

those suggests that Congress did not mean for the SCA’s protections to extend to

publicly configured posts. See S. Rep. 99-541, at 36 (“To access a communication

in such a public system is not a violation of the Act, since the general public has

been ‘authorized’ to do so.”). 47

The SCA is antiquated and could no doubt use a legislative update, but in the

meantime courts should read its provisions in a way that makes sense of the entire

statutory scheme, while cognizant of just how much has changed in the nearly-four

decades since it was passed. Doing that leads me to conclude that the SCA’s

protections do not extend to public posts, and the court should say so if a more

appropriate occasion ever arises.