Subscribe
Share
Share
Embed
This episode, Navigating the Digital Frontier: Regulation and limitations on Social Media features an expert panel that delve into the critical areas of law impacting social media platforms and users, including content moderation and platform liability, data privacy and security, free speech, and online expression
The panel features:
Alan Rozenshtein - Associate Professor of Law at Minnesota Law and senior editor and research director at Lawfare, a non-resident senior fellow at the Brookings Institution, and a term member of the Council on Foreign Relations.
Charlotte Garden - the Gray, Plant, Mooty, Mooty & Bennett Professor of Law at Minnesota Law; She specializes in labor law, employment law, and constitutional law.
The conversation is moderated by Jamie Snelson (‘97), Shareholder at Fredrikson & Byron and the co-chair of Minnesota Law’s Board of Advisors’ Academic Engagement Committee.
This event was recorded on April 21, 2025 as part of Spring Alumni Week. You can watch the entire event recording on the Minnesota Law YouTube channel. (https://youtu.be/XT1vcvzKx4M)
A transcript of this episode is available here: z.umn.edu/Ep52Transcript
Learn more about the University of Minnesota Law School by visiting law.umn.edu
LawTalk features events, webinars, and panel discussions hosted at the University of Minnesota Law School.
LawTalk Episode 52 - Navigating the Digital Frontier: Regulation and limitations on social media
Welcome to LawTalk, a podcast series produced by the University of Minnesota Law School featuring events, webinars, and panel discussions about diverse topics at the intersection of law, policy, and education.
This episode, Navigating the Digital Frontier: Regulation and limitations on Social Media features an expert panel that delve into the critical areas of law impacting social media platforms and users, including content moderation and platform liability, data privacy and security, free speech, and online expression
The panel features:
Alan Rozenshtein - Associate Professor of Law at Minnesota Law and senior editor and research director at Lawfare, a non-resident senior fellow at the Brookings Institution, and a term member of the Council on Foreign Relations.
Charlotte Garden - the Gray, Plant, Mooty, Mooty & Bennett Professor of Law at Minnesota Law; She specializes in labor law, employment law, and constitutional law.
The conversation is moderated by Jamie Snelson (‘97), Shareholder at Fredrikson & Byron and the co-chair of Minnesota Law’s Board of Advisors’ Academic Engagement Committee.
This event was recorded on April 21, 2025 as part of Spring Alumni Week. You can watch the entire event recording on the Minnesota Law YouTube channel. (https://youtu.be/XT1vcvzKx4M)
Subscribe to the Minnesota Law podcast feed on Apple Podcasts, Spotify, SoundCloud, or via your preferred podcast application by searching ‘University of Minnesota Law School’, for more LawTalk episodes, as well as other podcast content produced by Minnesota Law.
0:00:40.223,0:00:43.192
As you see, we have two members
of our illustrious
0:00:43.193,0:00:45.811
University of Minnesota
Law School faculty on hand
0:00:45.812,0:00:50.282
to explore some questions about
how social media is regulated
0:00:50.283,0:00:52.552
in general and then more specifically
0:00:52.553,0:00:55.137
how that regulation
might stifle speech,
0:00:55.138,0:00:58.741
how it might protect minors,
how it might impact workers' rights,
0:00:58.742,0:01:01.226
how it might allocate
risk among content
0:01:01.227,0:01:02.628
creators and platform owners
0:01:02.629,0:01:06.348
and maybe some equally
thorny issues as time allows.
0:01:06.349,0:01:11.036
With that as a little teaser,
let's meet our faculty experts.
0:01:11.037,0:01:13.822
Alan Rozenshtein is
an associate professor
0:01:13.823,0:01:17.631
at Minnesota Law,
where he has taught since 2017.
0:01:17.660,0:01:20.846
He's a senior editor
and research director at Lawfare,
0:01:20.847,0:01:22.347
a non-resident senior fellow
0:01:22.348,0:01:24.983
at the Brookings Institution,
and a term member
0:01:24.984,0:01:27.870
of the Council on Foreign Relations.
0:01:27.871,0:01:29.538
Before coming
to the University of Minnesota,
0:01:29.539,0:01:32.875
Professor Rozenshtein served
as an attorney advisor
0:01:32.876,0:01:35.961
in the Office of Law and Policy
in the National Security Division
0:01:35.962,0:01:38.147
of the US Department of Justice.
0:01:38.148,0:01:39.348
He's authored a wide range
0:01:39.349,0:01:41.183
of scholarship
on constitutional issues
0:01:41.184,0:01:44.516
and is a frequent
commentator in the press.
0:01:44.904,0:01:46.606
Professor Charlotte Garden
0:01:46.607,0:01:49.859
is the Gray, Plant, Mooty,
Mooty & Bennett Professor of Law
0:01:49.860,0:01:51.293
at Minnesota Law.
0:01:51.294,0:01:54.680
She joined the law
school in the fall of 2022.
0:01:54.681,0:01:57.649
She specializes in labor law,
employment law,
0:01:57.650,0:01:59.368
and constitutional law.
0:01:59.369,0:02:01.820
Her interests include
the intersection of workers' rights
0:02:01.821,0:02:04.456
and the Constitution,
how law supports
0:02:04.457,0:02:07.393
or undermines worker voice and power.
0:02:07.710,0:02:09.328
Prior to coming
to the University of Minnesota,
0:02:09.329,0:02:12.648
Professor Garden was a professor
at Seattle University School of Law,
0:02:12.649,0:02:16.318
and before that, she was a practicing
lawyer in Washington, DC.
0:02:16.319,0:02:18.455
Also, the author
of Impressive Scholarship,
0:02:18.789,0:02:21.023
much of it focused
on employment and labor issues,
0:02:21.024,0:02:24.293
including constitutional
issues in the labor environment.
0:02:24.294,0:02:26.829
To our panelists, again, welcome.
0:02:26.830,0:02:30.797
Thank you for sharing
your time and insights today.
0:02:30.984,0:02:33.886
Let's set the table with some basics.
0:02:33.887,0:02:37.039
Alan, maybe you could describe
for us the key aspects
0:02:37.040,0:02:40.108
of federal and state
regulatory frameworks
0:02:40.109,0:02:42.361
that are currently shaping
social media use
0:02:42.362,0:02:46.832
and how those frameworks
have evolved in the past few years.
0:02:46.833,0:02:47.833
Sure.
0:02:48.868,0:02:52.037
This is an easy question to answer
because there's not that much.
0:02:52.038,0:02:54.607
Congress is not very good
at regulating technology,
0:02:54.608,0:02:57.142
or maybe it is very good,
depending on what you think.
0:02:57.143,0:02:58.794
Maybe it is good
at regulating technology,
0:02:58.795,0:03:01.680
and that it does
not regulate technology.
0:03:01.681,0:03:04.733
With respect to social media,
there's not a ton.
0:03:04.734,0:03:06.602
Now, there is some stuff
about children,
0:03:06.603,0:03:09.688
which I'll get to,
and I can talk about it in a minute.
0:03:09.689,0:03:12.625
On the broad question
of social media,
0:03:13.309,0:03:14.393
very little, actually.
0:03:14.394,0:03:18.747
The main law is a law
that actually further deregulates
0:03:18.748,0:03:20.549
social media
from what you might think
0:03:20.550,0:03:22.217
of as the common law baseline.
0:03:22.218,0:03:24.953
That is Section 230,
which is this now infamous
0:03:24.954,0:03:31.526
or famous or both statute
that immunizes social media platforms
0:03:31.527,0:03:38.334
from harms caused by content
that is produced by their users.
0:03:38.935,0:03:41.336
That really has become the main piece
0:03:41.337,0:03:46.108
of legislation that has sort
of further empowered social media.
0:03:46.109,0:03:47.659
When it comes
to legislation that restricts
0:03:47.660,0:03:50.095
or regulates social media,
there's really basically nothing
0:03:50.096,0:03:54.266
to speak of at the federal
level with respect to adults.
0:03:54.267,0:03:56.868
At the state level,
things are somewhat different.
0:03:56.869,0:04:01.391
There are some laws that purport
to regulate social media.
0:04:02.200,0:04:03.893
In particular, California has
0:04:03.894,0:04:08.152
an important digital privacy
legislation that limits,
0:04:08.153,0:04:10.616
to some extent, what social media
0:04:10.617,0:04:12.718
can do with user information.
0:04:12.719,0:04:16.159
More notably,
Texas and Florida both have enacted
0:04:16.160,0:04:18.024
laws purporting to limit
0:04:18.025,0:04:21.198
what social media
companies can moderate.
0:04:22.228,0:04:24.479
They call these anti-censorship laws.
0:04:24.480,0:04:27.983
Both of the laws are mired
in litigation, in particular,
0:04:27.984,0:04:30.686
in the wake of the NetChoice decision
0:04:30.687,0:04:32.487
from the Supreme Court
from last year,
0:04:32.488,0:04:35.191
again, which we can talk about.
0:04:35.425,0:04:37.143
When it comes to children,
0:04:37.144,0:04:39.528
there you do see
some more regulation.
0:04:39.529,0:04:41.747
At the federal level,
the main piece of legislation
0:04:41.748,0:04:44.850
is the Children's Online
Privacy Protection Act, or COPPA.
0:04:44.851,0:04:50.156
This is from 1998, so this actually
predates modern social media.
0:04:50.390,0:04:56.778
This basically limits what internet
companies can do with respect
0:04:56.779,0:05:02.602
to the collection of data
on children 13 years and under.
0:05:03.236,0:05:05.838
It's not a meaningless statute,
0:05:06.589,0:05:08.190
but it's mostly a meaningless statute
0:05:08.191,0:05:13.345
because it does not really require
companies to do much
0:05:13.346,0:05:15.497
to actually prevent children
0:05:15.498,0:05:18.300
from under 13 from using
their services.
0:05:18.301,0:05:22.505
There are plenty of children
who are 12, 11, 10 years old
0:05:22.506,0:05:24.958
who lie about how old they are
0:05:24.975,0:05:26.658
and therefore get Instagram accounts
0:05:26.659,0:05:28.808
and things of that nature.
0:05:29.612,0:05:31.230
At the state level,
0:05:31.231,0:05:35.150
there are more muscular
child protection laws
0:05:35.151,0:05:36.451
that have been enacted,
0:05:36.452,0:05:40.973
are in the pipeline
in both blue and red states.
0:05:41.307,0:05:45.360
These are increasingly coming
into their own legal difficulty.
0:05:45.361,0:05:51.049
Texas, for example, passed
a law trying to put age verification
0:05:51.050,0:05:55.704
on pornographic sites,
making sure that people
0:05:55.705,0:05:58.825
under the age of 18 did
not access them.
0:05:59.609,0:06:02.477
That law is actually pending
before the Supreme Court right now.
0:06:02.478,0:06:06.081
Oral arguments
were held earlier this term.
0:06:06.082,0:06:09.234
There are other examples
of these kinds of legislation,
0:06:09.235,0:06:12.237
whether they're child safety laws
0:06:12.238,0:06:14.756
or whether they're child
verification laws.
0:06:14.757,0:06:17.392
Again, these are all ensnared
0:06:17.393,0:06:21.647
currently in a variety
of First Amendment litigation.
0:06:23.332,0:06:29.038
To give you an overall
on your question,
0:06:29.422,0:06:31.590
there's very little either
at the federal or state level
0:06:31.591,0:06:34.826
that is meaningfully
regulating social media,
0:06:34.827,0:06:38.897
and the one piece
of legislation that is important,
0:06:38.898,0:06:43.500
Section 230, is a highly deregulatory
piece of legislation.
0:06:45.037,0:06:50.842
You alluded a little bit,
I think, to other types of laws
0:06:50.843,0:06:54.564
that are in some cases
being challenged.
0:06:55.314,0:07:00.570
Are those the ones that relate
to content moderation and the like?
0:07:01.621,0:07:03.505
There are lots of things
that are being challenged.
0:07:03.506,0:07:06.475
Right now, the state content
moderation laws and again,
0:07:06.476,0:07:08.960
these are laws meant
to restrict content moderation,
0:07:08.961,0:07:12.497
so prevent platforms
from content moderating.
0:07:12.498,0:07:15.000
If you're a fan
of the bill censoring,
0:07:15.001,0:07:17.369
it depends on obviously
which side you're on.
0:07:17.370,0:07:20.272
Those are stuck
in post-NetChoice litigation.
0:07:20.273,0:07:22.374
You have child protection laws
0:07:22.375,0:07:24.709
which are not about
moderating content,
0:07:24.710,0:07:27.078
but they're about restricting
children from being able
0:07:27.079,0:07:31.349
to go on these websites
by requiring age verification,
0:07:31.350,0:07:34.769
often in the form
of the production of state ID.
0:07:34.770,0:07:37.522
That's the Texas pornography
law or anti-pornography law.
0:07:37.523,0:07:38.757
Those are being challenged, again,
0:07:38.758,0:07:39.691
not on the grounds that children
0:07:39.692,0:07:41.927
have a right to use
Texas pornography,
0:07:41.928,0:07:44.738
but on the grounds that adults do.
0:07:50.336,0:07:53.371
A law that even if aimed at children,
0:07:53.372,0:07:57.626
has the effect of overly burdening
adult access can be unconstitutional.
0:07:57.627,0:08:02.430
This was what the Supreme Court held
in ACLU v. Reno.
0:08:02.431,0:08:04.366
This is a landmark internet
0:08:04.367,0:08:06.085
First Amendment case
in the mid-1990s,
0:08:06.086,0:08:07.636
which struck down
0:08:07.637,0:08:09.204
what was then Congress's main attempt
0:08:09.205,0:08:10.139
to regulate the internet,
0:08:10.140,0:08:12.090
which is the Communications
Decency Act.
0:08:12.091,0:08:15.210
It struck down a provision
that would criminalize
0:08:15.211,0:08:18.180
giving children access
to pornography again,
0:08:18.181,0:08:21.233
on the grounds that have
an excessively chilling effect
0:08:21.234,0:08:24.119
on the ability of adults
to access that kind of material.
0:08:24.120,0:08:27.928
Then there are other laws
that directly require--
0:08:28.375,0:08:30.692
I think there's a law
in California and a law in New York,
0:08:30.693,0:08:32.210
the law in New York might be pending,
0:08:32.211,0:08:34.079
I don't know
if it's been enacted yet,
0:08:34.080,0:08:37.699
that would require
social media platforms
0:08:37.700,0:08:44.422
to modify their services
in the best interest of children.
0:08:44.423,0:08:47.042
Those laws are tied
up in litigation on void
0:08:47.043,0:08:49.194
for vagueness
First Amendment grounds.
0:08:49.195,0:08:51.646
The First Amendment
is a pretty big buzzsaw.
0:08:51.647,0:08:54.766
For this, I will say, though,
we're still very, I think,
0:08:54.767,0:08:58.103
early in what will be plausibly
half a decade,
0:08:58.104,0:09:00.506
if not longer, of litigation.
0:09:01.140,0:09:04.643
We'll see what the First Amendment
looks like at the end of this.
0:09:04.644,0:09:06.745
Right now, the case law
is pretty libertarian.
0:09:06.746,0:09:08.430
It's pretty deregulatory.
0:09:08.431,0:09:11.783
I think lower courts
are correctly viewing
0:09:11.784,0:09:14.452
a lot of these laws
with doctrinal skepticism,
0:09:14.453,0:09:18.106
even though I think it's very much
an open question how much
0:09:18.107,0:09:20.458
the Supreme Court might ultimately,
0:09:20.459,0:09:23.995
again, with respect to children,
allow for more regulation.
0:09:23.996,0:09:27.499
Again, that's going to take us five,
six, seven years to figure out.
0:09:27.500,0:09:28.701
Understood.
0:09:29.151,0:09:30.752
Let's switch gears a little bit.
0:09:30.753,0:09:34.561
We can come back
to any of that as you each like.
0:09:34.707,0:09:37.676
Charlotte, maybe
you could talk a little bit about
0:09:37.677,0:09:41.963
how the evolution of social media
platforms and their use by employees
0:09:41.964,0:09:46.284
and employers has impacted
traditional labor law concepts
0:09:46.285,0:09:48.054
and practices, too.
0:09:48.120,0:09:49.120
Yes, sure.
0:09:51.107,0:09:56.579
One thing I will say
is that that question,
0:09:56.696,0:09:58.980
there's more than a decade
of the board dealing
0:09:58.981,0:10:01.016
with the application
0:10:01.017,0:10:05.063
of the National Labor
Relations Act to social media.
0:10:06.589,0:10:07.623
There are some questions
0:10:07.624,0:10:12.294
that are pretty
well answered, such as,
0:10:12.295,0:10:15.947
can protected concerted
activity take place online?
0:10:15.948,0:10:17.900
Resoundingly, yes.
0:10:19.652,0:10:22.671
There, it's not
a question of it being
0:10:22.672,0:10:24.239
really difficult to translate
0:10:24.240,0:10:28.076
an in-real-life concept
to the internet or to social media.
0:10:28.077,0:10:31.830
It's just to the extent workers
0:10:31.831,0:10:33.632
can talk about their working
conditions at work,
0:10:33.633,0:10:35.867
they can also talk about
them on social media.
0:10:35.868,0:10:38.771
There were some hiccups
early on about
0:10:38.772,0:10:41.723
if somebody likes a post,
is that the same?
0:10:41.724,0:10:45.176
Does that qualify as evidence towards
0:10:45.177,0:10:49.382
the concertedness
requirement having being met?
0:10:50.333,0:10:52.168
Yes, it turns out.
0:10:53.519,0:10:56.571
That question
has a really close parallel,
0:10:56.572,0:10:58.440
the question
that can arise in real life.
0:10:58.441,0:11:01.409
If somebody tries to initiate
protected concerted activity,
0:11:01.410,0:11:04.462
it doesn't really work,
can that be covered
0:11:04.463,0:11:06.297
by the National Labor Relations Act?
0:11:06.298,0:11:07.978
Again, often, yes.
0:11:09.001,0:11:13.555
Likewise, employers making coercive
threats about their employees,
0:11:13.556,0:11:16.975
that can happen
both in real life and online,
0:11:16.976,0:11:23.581
and courts will treat
the two contexts fairly similarly,
0:11:23.582,0:11:25.767
or the board will,
and courts will affirm.
0:11:25.768,0:11:28.887
There are a couple
of recent high-profile cases
0:11:28.888,0:11:33.258
involving employers'
arguably coercive threats
0:11:33.259,0:11:36.811
on social media
and their treatment by the NLRB.
0:11:36.812,0:11:38.780
I'll maybe just mention them.
0:11:38.781,0:11:42.193
They're both just
intrinsically interesting.
0:11:43.436,0:11:45.353
They both involved tweets,
0:11:45.354,0:11:48.106
but one involved
a tweet by Dave Portnoy,
0:11:48.107,0:11:51.077
the Barstool Sports guy.
0:11:51.293,0:11:56.281
He tweeted something like,
"The first of you,"
0:11:56.282,0:11:58.083
you being his employees,
"The first of you
0:11:58.084,0:12:01.336
to unionize gets sent
back to the salt mines."
0:12:01.337,0:12:05.190
The NLRB ended up saying
that's a coercive threat.
0:12:05.191,0:12:08.682
Third Circuit reversed,
saying that's a joke.
0:12:09.512,0:12:12.498
The question is really,
how do you deal with
0:12:12.499,0:12:15.333
or how do you figure
out context online?
0:12:15.334,0:12:17.648
Would an employee be coerced
0:12:18.320,0:12:23.909
in their ability to organize a union,
or would they not?
0:12:25.261,0:12:27.595
The other case involves Elon Musk.
0:12:27.596,0:12:30.133
He tweeted ages ago
0:12:31.017,0:12:36.989
that if his Tesla employees
unionized, they--
0:12:36.990,0:12:39.441
it was something like,
"Why would you unionize
0:12:39.442,0:12:42.710
when it's just going to mean
that you lose your stock options?"
0:12:42.711,0:12:46.582
The NLRB again said
that was a coercive threat.
0:12:47.099,0:12:50.668
The Fifth Circuit recently
heard that case, and they,
0:12:50.669,0:12:54.122
in what I would say
is like a really strange opinion,
0:12:54.123,0:12:58.376
said, "We think that it's fine
that the board did that,
0:12:58.377,0:13:00.712
but we also think
the First Amendment prohibits
0:13:00.713,0:13:04.849
the board from ordering
that the tweet be taken down."
0:13:04.850,0:13:05.900
The Fifth Circuit's view
0:13:05.901,0:13:10.071
is that somebody
can host on social media
0:13:10.072,0:13:13.892
a threat that violates labor
law because it's coercive,
0:13:13.893,0:13:18.098
but that the board is powerless
to have it be removed.
0:13:18.681,0:13:20.098
That is an interesting view
0:13:20.099,0:13:23.828
of the scope of the First
Amendment, I will say.
0:13:23.936,0:13:27.055
I tend to think
that it's one way or the other.
0:13:27.056,0:13:29.724
Either it's protected speech
because it's non-coercive,
0:13:29.725,0:13:32.911
and that's also why it doesn't
violate labor law, or the reverse.
0:13:32.912,0:13:35.280
It violates labor law
because it's coercive
0:13:35.281,0:13:37.398
and therefore can be ordered
to be taken down.
0:13:37.399,0:13:42.239
I find the idea that something
can be both to be very unusual.
0:13:42.955,0:13:49.277
There are some other emerging
issues related to, for example,
0:13:49.278,0:13:51.563
employers' surveillance
of social media activity,
0:13:51.564,0:13:53.431
which I think we'll talk
about a little bit later,
0:13:53.432,0:13:55.650
but I think I've been talking
for a while now,
0:13:55.651,0:13:58.296
so let me hand it back to Jamie.
0:14:01.140,0:14:06.661
You both telegraphed that emerging
issues are being addressed
0:14:06.662,0:14:10.532
and heard, and hopefully
decided by the courts.
0:14:10.533,0:14:13.017
What's your prediction here as we sit
0:14:13.018,0:14:16.554
in a dynamic point in 2025 about
0:14:16.555,0:14:21.776
what's going to come
up first and be decided,
0:14:21.777,0:14:24.329
or where the issues or resolution
0:14:24.330,0:14:27.665
of them will be advanced
as we look forward here
0:14:27.666,0:14:31.520
in 2025 and maybe into 2026 too?
0:14:32.087,0:14:34.236
Open question to you both.
0:14:36.859,0:14:38.694
I'm happy to start.
0:14:40.529,0:14:41.818
I'm not sure.
0:14:42.948,0:14:49.220
The Texas pornography child
verification case is pending,
0:14:49.221,0:14:51.256
but the procedural
posture of that one is weird
0:14:51.257,0:14:55.410
because the Fifth Circuit,
as it has been doing lately,
0:14:55.411,0:14:59.697
played very weirdly fast and loose
with the controlling precedent
0:14:59.698,0:15:01.849
and read it in such
a way as to apply,
0:15:01.850,0:15:05.453
I think, no scrutiny at all,
which is an interesting move.
0:15:05.454,0:15:07.889
I suspect what the court
will do is will just remand
0:15:07.890,0:15:09.857
to apply the correct standard,
0:15:09.858,0:15:12.443
and then will just be back
at the races of intermediate
0:15:12.444,0:15:15.563
or strict scrutiny,
or intermediate plus or strict minus,
0:15:15.564,0:15:19.685
and the scholastic
head-on-a-pin nature of a lot
0:15:19.686,0:15:24.657
of this First Amendment
doctrinal analysis.
0:15:24.890,0:15:27.475
Again, the same thing with NetChoice.
0:15:27.476,0:15:32.513
We're stuck in this lower
court endless litigation
0:15:32.514,0:15:34.465
over facial
versus as-applied challenges,
0:15:34.466,0:15:36.551
and how do you do
the as-applied challenge?
0:15:36.552,0:15:38.369
That'll take a long time to get back.
0:15:38.370,0:15:40.438
I think what's fun
about this area of law,
0:15:40.439,0:15:43.341
and I'm very curious, Charlotte,
if you agree with me on this,
0:15:43.342,0:15:47.662
is that maybe the labor side
of this is somewhat different,
0:15:47.663,0:15:51.154
but at least for the pure
social media cases,
0:15:51.717,0:15:54.919
there's not an obvious
political valence to a lot of this.
0:15:54.920,0:15:58.823
This isn't like abortion or gun
rights or affirmative action,
0:15:58.824,0:16:01.309
where you really know
0:16:01.310,0:16:03.294
what every single
justice is going to say,
0:16:03.295,0:16:07.949
and the oral argument has this kabuki
theater-like nature to it.
0:16:07.950,0:16:10.018
Again, not because
anyone's acting dishonestly,
0:16:10.019,0:16:11.903
but it's really baked in.
0:16:11.904,0:16:15.990
Here you have these really
interesting cross coalitions
0:16:15.991,0:16:17.759
that just don't seem that related
0:16:17.760,0:16:21.295
to the underlying policy
or political preferences
0:16:21.296,0:16:22.463
of the justices, and you really saw
0:16:22.464,0:16:25.583
that with NetChoice
where you had Kagan writing
0:16:25.584,0:16:28.186
an opinion for her and Sotomayor
0:16:28.187,0:16:29.821
and the chief and Kavanaugh
0:16:29.822,0:16:33.358
and Barrett and Jackson, but Jackson,
0:16:33.359,0:16:34.960
only signed
on to some of that opinion,
0:16:34.961,0:16:38.246
and Barrett wrote a concurrence
that undermined a lot of the opinion,
0:16:38.247,0:16:42.533
and then you had Alito and Thomas
and Gorsuch on the other side.
0:16:42.534,0:16:47.673
I think you could see
something very similar
0:16:55.347,0:16:56.431
with the Texas case.
0:16:56.432,0:16:57.899
You saw something
that you haven't talked about,
0:16:57.900,0:16:59.600
but maybe worth
talking about a little bit.
0:16:59.601,0:17:02.270
We saw that a little bit
with the TikTok opinion,
0:17:02.271,0:17:04.756
which was a 9-0 opinion
upholding the TikTok ban,
0:17:04.757,0:17:08.143
but the two justices who clearly
were least happy about it
0:17:08.144,0:17:11.030
were Sotomayor and Gorsuch.
0:17:13.232,0:17:15.133
One of the things
that makes this really fun
0:17:15.134,0:17:19.755
as a scholar
is the unpredictable nature of it
0:17:19.772,0:17:22.290
and the sense that the justices,
0:17:22.291,0:17:25.493
they're really thinking
from first principles rather than,
0:17:25.494,0:17:28.346
"Well, I thought
X on this for the last 20 years,
0:17:28.347,0:17:32.314
so that's what I'm going to write
on this opinion."
0:17:33.268,0:17:34.469
I really agree with that.
0:17:34.470,0:17:36.587
I think one of the really interesting
things about
0:17:36.588,0:17:41.876
the NetChoice decision
was the way some justices,
0:17:41.877,0:17:44.846
both justices more on the left
and justices more on the right,
0:17:44.847,0:17:48.868
flipped their usual
position on the extent
0:17:48.901,0:17:53.171
to which the government
can try to create
0:17:53.172,0:17:57.392
a more fair marketplace for speech.
0:17:58.794,0:18:00.962
Jamie, let us know
if you want us to--
0:18:00.963,0:18:01.913
I don't know.
0:18:01.914,0:18:03.498
I don't want to steal
your moderator thunder,
0:18:03.499,0:18:05.716
but if you want
to get into NetChoice,
0:18:05.717,0:18:06.835
we can give a 30-second
0:18:06.836,0:18:09.087
on what it is we're talking about.
0:18:09.088,0:18:11.272
These are content moderation cases.
0:18:11.273,0:18:14.342
If you could give us the quick sip,
that would be great.
0:18:14.343,0:18:15.694
Why don't I give the 30-second
0:18:15.695,0:18:17.379
and then go back to you, Charlotte-
0:18:17.380,0:18:18.313
Sure.
0:18:18.314,0:18:20.297
-on what the opinion is.
0:18:20.432,0:18:23.668
The NetChoice cases
arose from two state laws,
0:18:23.669,0:18:25.937
one out of Florida,
one out of Texas, that each tried
0:18:25.938,0:18:29.307
to limit what social media
companies could moderate.
0:18:29.308,0:18:32.777
The background to this is that,
obviously, platforms moderate.
0:18:32.778,0:18:35.897
Moderation is, in some sense,
the product, actually.
0:18:35.898,0:18:37.386
It's very core.
0:18:37.683,0:18:40.852
There has been, for a long time,
a concern among conservatives
0:18:40.853,0:18:44.839
that social media companies
are biased against them.
0:18:44.840,0:18:47.567
The reality is quite complicated.
0:18:48.360,0:18:50.328
It's not clear that they're biased
against them.
0:18:50.329,0:18:52.697
They might be biased against them.
0:18:52.698,0:18:56.100
Certainly, the people
at the platform's code left,
0:18:56.101,0:18:59.337
though maybe that changed
in the last six months.
0:18:59.338,0:19:01.105
The reality is very complicated.
0:19:01.106,0:19:03.357
There's certainly been a perception
that this is the case.
0:19:03.358,0:19:04.458
The Texas and Florida laws
0:19:04.459,0:19:06.928
were actually
very self-consciously written,
0:19:06.929,0:19:10.658
pushed back against
this perceived liberal bias.
0:19:16.038,0:19:17.905
The problem and the opportunity
for the laws, actually,
0:19:17.906,0:19:20.908
is that there's a very complicated
doctrine going all the way back
0:19:20.909,0:19:25.364
to the 19-- even before,
but the '60s and 1970s,
0:19:26.165,0:19:31.102
about when the government
can require individuals
0:19:31.103,0:19:34.856
and platforms and companies
to host other people's speech.
0:19:34.857,0:19:38.359
There's this one set of cases
going all the way back
0:19:38.360,0:19:41.363
to this famous case called
Miami Herald v. Tornillo,
0:19:41.364,0:19:42.780
in which the court says, look,
0:19:42.781,0:19:45.417
companies have First
Amendment rights,
0:19:45.418,0:19:49.020
editorial rights
as to what they themselves carry.
0:19:49.021,0:19:52.707
Then there's a different set
of cases, cases like Pruneyard,
0:19:52.708,0:19:55.193
cases like,
now I'm embarrassed to say,
0:19:55.194,0:19:57.712
I forget the name of the case,
but involving actually law schools
0:19:57.713,0:20:00.831
and whether they can
be forced to host
0:20:00.832,0:20:04.018
military recruiters,
where in other circumstances,
0:20:04.019,0:20:06.571
the court says, "Look,
there are certain circumstances
0:20:06.572,0:20:08.189
where hosting someone else's speech
0:20:08.190,0:20:09.407
is not an expressive activity,
0:20:09.408,0:20:11.609
so the First Amendment
doesn't really apply."
0:20:11.610,0:20:14.729
The question for this case was,
0:20:14.730,0:20:16.347
what are social media companies like?
0:20:16.348,0:20:21.018
Are they more like newspapers
0:20:21.019,0:20:23.187
which have expressive
editorial interests,
0:20:23.188,0:20:25.823
or are they more like shopping
malls and law schools
0:20:25.824,0:20:28.877
that don't have expressive interests?
0:20:29.428,0:20:32.146
Simultaneously, the other wrinkle
0:20:32.147,0:20:34.165
for all of this was
that when NetChoice,
0:20:34.166,0:20:35.850
which is the industry organization
0:20:35.851,0:20:37.285
that brought these cases,
brought these cases,
0:20:37.286,0:20:38.753
they brought them
as facial challenges.
0:20:38.754,0:20:40.321
They said, "Look, these laws
0:20:40.322,0:20:43.496
are always and forever
unconstitutional."
0:20:44.509,0:20:46.761
The courts, the 11th Circuit,
which struck down the law,
0:20:46.762,0:20:48.195
and the Fifth Circuit,
which upheld the law,
0:20:48.196,0:20:51.199
treated the litigation as facial.
0:20:51.350,0:20:53.034
The problem is,
it got up to the Supreme Court,
0:20:53.035,0:20:54.819
and very quickly it became clear
0:20:54.820,0:20:57.838
that these were not
good facial lawsuits,
0:20:57.839,0:21:02.927
because actually the breadth
of these laws was quite stark,
0:21:02.928,0:21:08.182
and they applied not just
to the paradigmatic example,
0:21:08.183,0:21:10.918
which is, let's say,
the public Facebook newsfeed,
0:21:10.919,0:21:15.656
but to a WhatsApp private message,
or to Uber Eats,
0:21:15.657,0:21:18.219
or to a million other examples.
0:21:18.293,0:21:21.562
They might also apply
to AI algorithmic moderation,
0:21:21.563,0:21:23.581
and that might be its own thing.
0:21:23.582,0:21:26.835
Basically, what the opinion
in NetChoice did
0:21:26.836,0:21:28.636
was maybe punted as unfair,
0:21:28.637,0:21:29.670
because this really was a problem.
0:21:29.671,0:21:32.690
It said, "Look,
these are not good facial challenges.
0:21:32.691,0:21:34.592
Everyone needs to start
over from the beginning
0:21:34.593,0:21:38.614
and do them as applied,
zeroing in on exactly
0:21:38.631,0:21:41.148
what the specific litigant
is complaining about
0:21:41.149,0:21:42.566
with the law, with respect to them,
0:21:42.567,0:21:44.819
and then we will try this again."
0:21:44.820,0:21:46.370
In addition to doing that,
0:21:46.371,0:21:49.557
the various coalitions
on the court also wrote
0:21:49.558,0:21:52.259
a bunch of dicta saying,
"When you do that,
0:21:52.260,0:21:53.928
keep these things in mind."
0:21:53.929,0:21:58.015
The Kagan majority wrote
some dicta that was very strong
0:21:58.016,0:22:01.666
on the First Amendment rights
of the platforms.
0:22:01.887,0:22:04.105
Justice Barrett split
the baby a little bit,
0:22:04.106,0:22:07.792
saying, "Generally, yes, but AI
algorithms might be different."
0:22:07.793,0:22:09.710
The Alito opinion basically said,
0:22:09.711,0:22:11.512
"This isn't speech,
this is censorship."
0:22:11.513,0:22:13.798
The Jackson opinion, I think
very much to her credit, said,
0:22:13.799,0:22:15.116
"Why are we doing all this dicta?
0:22:15.117,0:22:15.916
Dicta is bad.
0:22:15.917,0:22:19.787
I don't want to do this until
I have to in five years."
0:22:19.788,0:22:21.856
That's why we're back
in this situation.
0:22:21.857,0:22:26.311
Sorry for that mini lecture,
but Charlotte, back to you.
0:22:28.497,0:22:32.717
One thing I will say
is that since the NetChoice decision,
0:22:32.718,0:22:34.585
the case has been remanded
to the Fifth Circuit,
0:22:34.586,0:22:37.288
and then from there,
remanded to the district courts.
0:22:37.289,0:22:39.642
The Fifth Circuit said
0:22:40.042,0:22:43.057
there has to be just
so much discovery.
0:22:43.428,0:22:46.464
Alan, I think you said
there could be litigation
0:22:46.465,0:22:48.232
over these issues for half a decade.
0:22:48.233,0:22:50.235
I think there could be discovery
in this case
0:22:50.236,0:22:54.438
for half a decade
because the NetChoice
0:22:54.439,0:22:59.060
plaintiffs tried to say, "On remand,
we should be allowed
0:22:59.061,0:23:03.531
to develop these factual
issues to a limited degree."
0:23:03.532,0:23:06.016
The court was like, "Oh no,
you have to go back
0:23:06.017,0:23:09.954
and figure out how every platform
0:23:09.955,0:23:14.275
to which this law might apply
engages in content moderation
0:23:14.276,0:23:17.928
and the extent to which that content
moderation qualifies as expressive
0:23:17.929,0:23:20.131
and is therefore covered
by the First Amendment.
0:23:20.132,0:23:24.018
What we have to do is ultimately
be able to say,
0:23:24.019,0:23:27.054
'Here's the scope of situations
to which this law applies
0:23:27.055,0:23:29.039
that raise a First Amendment problem,
0:23:29.040,0:23:31.258
and here's the scope of situations
to which this law applies
0:23:31.259,0:23:33.110
that does not raise
a First Amendment problem.'
0:23:33.111,0:23:34.428
We have to be able to compare them,
0:23:34.429,0:23:35.696
and that means we have to know
0:23:35.697,0:23:41.253
the precise weight of both sides
of the scale," if you like.
0:23:43.472,0:23:46.340
I'm sure NetChoice
was deeply unthrilled
0:23:46.341,0:23:51.815
to get that order and get sent back
for extremely extensive discovery.
0:23:55.133,0:24:00.688
In the Kagan opinion, in his dicta,
there is some discussion
0:24:00.689,0:24:03.858
of whether and when
the state's interests
0:24:03.859,0:24:05.977
in regulating content moderation
0:24:05.978,0:24:09.748
might satisfy heightened
scrutiny that applies.
0:24:11.183,0:24:15.836
The Kagan majority says
the government can't just assert
0:24:15.837,0:24:20.691
an interest in improving or balancing
the marketplace of ideas.
0:24:20.692,0:24:22.372
That is interesting.
0:24:22.494,0:24:24.962
The Kagan majority
says this is related
0:24:24.963,0:24:26.397
to the suppression of free ideas,
0:24:26.398,0:24:30.524
it just cannot be a qualifying
state interest at all.
0:24:34.958,0:24:38.209
That is consistent with various
important First Amendment cases,
0:24:38.210,0:24:40.044
including Citizens United.
0:24:40.045,0:24:42.947
In Citizens United, the court says
0:24:42.948,0:24:45.599
that what in the election
context is called
0:24:45.600,0:24:47.997
the anti-distortion rationale
0:24:51.173,0:24:53.057
can't qualify
as a government interest
0:24:53.058,0:24:56.694
because it inherently
violates the First Amendment.
0:24:56.695,0:25:01.398
When in previous cases,
the court had said government
0:25:01.399,0:25:04.068
can regulate corporate
election spending in order
0:25:04.069,0:25:07.221
to achieve a fairer
marketplace of ideas.
0:25:07.222,0:25:08.610
The court says
0:25:13.011,0:25:15.430
that's trying to affect
the composition
0:25:15.431,0:25:17.332
of people's expression
0:25:17.716,0:25:19.667
or of the total
marketplace of expression
0:25:19.668,0:25:22.970
in order to change its composition.
0:25:22.971,0:25:25.706
That's inherently
a First Amendment problem.
0:25:25.707,0:25:29.693
In the election context,
it's been the more liberal justices
0:25:29.694,0:25:31.612
that say, government ought
0:25:31.613,0:25:35.399
to be able to regulate
in a way that is aimed
0:25:35.400,0:25:40.160
at avoiding market distortion
by just people with more money.
0:25:40.222,0:25:42.106
The more conservative
justices saying, "No,
0:25:42.107,0:25:44.008
that is inherently
a First Amendment problem."
0:25:44.009,0:25:45.976
In this case, you see that reversed
0:25:45.977,0:25:49.114
and just do a nearly perfect 180
0:25:49.615,0:25:52.199
with the more liberal
justices saying,
0:25:52.200,0:25:55.436
"No, that's categorically
not permissible."
0:25:55.437,0:26:02.159
In a way, I don't read it as,
we've said this is impermissible.
0:26:02.160,0:26:04.078
I didn't agree with it there,
but it's the rule.
0:26:04.079,0:26:06.230
I read it as a more
full-throat endorsement
0:26:06.231,0:26:09.316
of the idea that
the anti-distortion principle
0:26:09.317,0:26:12.253
is just unacceptable in this context.
0:26:12.737,0:26:14.004
I think that's interesting.
0:26:14.005,0:26:16.140
I think it reflects a real shakeup.
0:26:16.141,0:26:20.277
Both of the laws here
are very much conservative coded,
0:26:20.278,0:26:24.866
but as Alan pointed out,
there are California, New York,
0:26:25.133,0:26:26.750
other states passing other kinds
0:26:26.751,0:26:31.489
of content moderation laws
that would code the other way.
0:26:31.856,0:26:33.173
Can I just say something,
really, just to jump
0:26:33.174,0:26:36.027
on something Charlotte said?
0:26:36.661,0:26:38.178
On this anti-distortion rationale,
0:26:38.179,0:26:40.648
one thing that I was surprised
about and thought was somewhat
0:26:40.649,0:26:44.568
short-sighted is there were a lot
of progressive
0:26:44.569,0:26:49.039
legal types who very much
disliked the Texas and Florida laws.
0:26:49.040,0:26:51.892
I understand why they did,
they weren't great laws.
0:26:51.893,0:26:54.228
I don't think they appreciated
0:26:54.229,0:26:57.531
that this is an opportunity
for the court to cut
0:26:57.532,0:27:02.220
back on this resistance
to anti-distortion arguments.
0:27:02.370,0:27:05.339
It is exactly this resistance
to anti-distortion arguments
0:27:05.340,0:27:06.908
that has led to a regime
0:27:06.909,0:27:09.694
of this very Lochnerized
First Amendment,
0:27:09.695,0:27:12.647
as the academics say,
that progressives,
0:27:12.648,0:27:17.969
I think with some justification,
really dislike.
0:27:19.437,0:27:20.904
For folks following along,
0:27:20.905,0:27:24.342
that's the two NetChoice cases
0:27:24.960,0:27:27.328
that Charlotte and Alan
have been talking about.
0:27:27.329,0:27:30.615
Moody v. NetChoice
and NetChoice v. Paxton
0:27:30.616,0:27:34.586
are the two cases
and both were amended, correct?
0:27:35.186,0:27:36.388
Yes.
0:27:38.073,0:27:41.875
Moving along a little bit, Alan,
I think you mentioned TikTok,
0:27:41.876,0:27:42.960
and I want to come back to that.
0:27:42.961,0:27:45.179
Before that,
let's talk a little bit about AI,
0:27:45.180,0:27:49.166
which I think you brought
into the discussion, Charlotte.
0:27:49.167,0:27:52.186
Seems like generative
AI in particular
0:27:52.187,0:27:55.916
might be a transformative
force on social media,
0:27:56.208,0:27:58.726
heightening risks like misinformation
0:27:58.727,0:28:01.627
and manipulation, maybe many others.
0:28:01.680,0:28:02.697
Is that true?
0:28:02.698,0:28:05.282
If so, does that underscore
the urgency for regulation
0:28:05.283,0:28:08.553
or just make meaningful
regulation harder?
0:28:09.921,0:28:10.338
Any thoughts [crosstalk]
0:28:10.338,0:28:11.423
Let's see.
0:28:11.790,0:28:13.474
I think this question
is cued up to let me talk
0:28:13.475,0:28:17.612
about my imminently
forthcoming edited volume.
0:28:18.446,0:28:19.130
Let's see.
0:28:19.131,0:28:20.097
I don't know
if that's what Jamie meant,
0:28:20.098,0:28:21.565
but that's what I'm going to do.
0:28:21.566,0:28:26.820
[chuckles] I am a co-editor
of a forthcoming volume,
0:28:26.821,0:28:30.024
should be hitting
shelves any day now,
0:28:30.175,0:28:33.983
on misinformation,
disinformation, and democracy,
0:28:34.229,0:28:37.181
where we have a number
of contributors
0:28:37.182,0:28:42.086
from a range of countries,
0:28:42.087,0:28:45.139
talking about a range
of countries and regions,
0:28:45.140,0:28:46.490
and their various approaches
0:28:46.491,0:28:51.262
to regulating mis
and disinformation on social media.
0:28:52.347,0:28:55.649
When we started this project,
I would say that in particular,
0:28:55.650,0:28:58.786
one of my co-editors thought,
"Oh, maybe we're going to find
0:28:58.787,0:29:02.489
the country that has figured
this out, and has figured out
0:29:02.490,0:29:04.775
how to meaningfully
deal with the problem
0:29:04.776,0:29:07.173
of online mis-disinformation,
0:29:07.529,0:29:11.331
while preserving
the robust free speech rights
0:29:11.332,0:29:14.635
that ought to be part
of a liberal democracy."
0:29:14.636,0:29:17.968
As it turns out,
that is not what happened.
0:29:18.857,0:29:23.627
The contributors to our book
explain a range of balances
0:29:23.628,0:29:28.230
that the countries or regions
they write about have struck,
0:29:28.683,0:29:35.022
balancing on one hand,
free speech against the harms,
0:29:35.023,0:29:37.074
including electoral
or political harms
0:29:37.075,0:29:40.011
that come with mis
and disinformation.
0:29:40.311,0:29:44.148
I think, like the phrase,
the devil is in the details,
0:29:44.149,0:29:47.049
it may appear in like every chapter.
0:29:47.268,0:29:53.240
The upshot is that, mainly,
0:29:53.241,0:29:54.926
to the extent countries
are really trying to deal with this,
0:29:54.927,0:29:57.861
often through
a self-regulatory mechanism,
0:29:57.862,0:29:59.446
like putting pressure on platforms
0:29:59.447,0:30:03.000
to do more and report
what they're doing
0:30:03.001,0:30:05.402
and report how effective
it is to respond to miss
0:30:05.403,0:30:07.938
and disinformations
like the EU approach.
0:30:07.939,0:30:10.140
Really tries to encourage platforms
0:30:10.141,0:30:14.584
to do things like demonetize
disinformation distributors.
0:30:15.530,0:30:19.650
As far as I can tell, the subsequent
research mostly suggested
0:30:19.651,0:30:22.745
that hasn't particularly
been effective.
0:30:24.806,0:30:27.708
I would say that this is a problem
that we have not solved
0:30:27.709,0:30:30.028
and is perhaps unsolvable.
0:30:31.563,0:30:34.816
If I may just hop
on to that for a second,
0:30:34.999,0:30:36.867
generative AI, and maybe I'm biased
0:30:36.868,0:30:40.070
because this is essentially
all I study now,
0:30:40.071,0:30:43.191
is such a tectonic technology.
0:30:45.510,0:30:49.229
At least I think that it is such
a huge deal that it's never
0:30:49.230,0:30:52.562
straightforwardly good
or bad for anything.
0:30:53.268,0:30:56.186
I think it could be a disaster
for mis and disinformation.
0:30:56.187,0:30:59.607
It could be really good
as well, because frankly,
0:30:59.608,0:31:03.810
ChatGPT is a lot more rational
and reality-based than most people,
0:31:03.811,0:31:06.704
myself included, a lot of the time.
0:31:09.017,0:31:10.334
One can imagine
a world in which people
0:31:10.335,0:31:12.085
actually get a lot better information
0:31:12.086,0:31:15.006
and a lot more calm
and rational information
0:31:15.007,0:31:18.091
by talking to these chatbots
in some way.
0:31:18.092,0:31:21.061
Also, one can imagine
a tuned version of this,
0:31:21.062,0:31:22.997
some bad actor takes
an open source model
0:31:22.998,0:31:25.315
and tunes it to be as inflammatory
0:31:25.316,0:31:28.636
and as disinformatory as possible,
0:31:28.637,0:31:32.756
and then uses video
and image generators to create
0:31:32.757,0:31:37.110
an army of bots and utterly
destroys the internet.
0:31:37.111,0:31:41.181
One can imagine both of those things
happening at the same time.
0:31:41.182,0:31:44.101
There's this great phrase,
it's the devil's in the details,
0:31:44.102,0:31:47.839
that I think applies here as well.
0:31:48.139,0:31:49.257
Understood.
0:31:49.490,0:31:50.991
I mentioned it a minute ago,
0:31:50.992,0:31:53.393
so I'll go back
to it at least for a moment.
0:31:53.394,0:31:57.798
When we first scheduled this session,
we all thought
0:31:57.799,0:32:00.450
that there would be big
TikTok news to share,
0:32:00.451,0:32:02.352
and there really isn't.
0:32:02.754,0:32:04.071
Curious to get your thoughts
0:32:04.072,0:32:07.791
on where that's likely
to go in this period,
0:32:07.792,0:32:09.293
where we're still waiting
0:32:09.294,0:32:12.930
what's going to happen
to TikTok in the United States.
0:32:12.931,0:32:16.422
For my sins,
I have somehow become TikTok guy
0:32:16.434,0:32:20.270
because I've been following
this obsessively for the last year.
0:32:20.271,0:32:23.757
I have to say,
it's become a real farce
0:32:23.758,0:32:25.176
because you have a situation
0:32:25.177,0:32:28.161
in which you have a law
that is extremely unambiguous.
0:32:28.162,0:32:32.516
It was upheld by nine justices
and three lower court judges.
0:32:32.517,0:32:35.486
We got 12-0.
0:32:36.321,0:32:38.387
The Trump administration,
0:32:38.507,0:32:40.407
and I should say
that I put some blame
0:32:40.408,0:32:43.544
on congressional Democrats
and also the Biden administration
0:32:43.545,0:32:45.562
for also going soft
and wobbly on this.
0:32:45.563,0:32:46.980
In particular,
the Trump administration,
0:32:46.981,0:32:50.017
which last time I checked,
has a duty to take care
0:32:50.018,0:32:53.420
that the laws be faithfully executed,
is just not enforcing
0:32:53.421,0:32:58.825
the law and is inventing totally
made-up extensions based
0:32:58.826,0:33:01.312
on nothing, truly nothing.
0:33:01.879,0:33:05.666
Is convincing the companies
that do business with TikTok,
0:33:05.667,0:33:08.969
Oracle and Google,
and Apple to go along with this,
0:33:08.970,0:33:12.556
despite the fact that the laws
penalties for doing business
0:33:12.557,0:33:14.808
with TikTok are actually quite high.
0:33:14.809,0:33:16.259
Especially if you're
Oracle and you're providing
0:33:16.260,0:33:19.696
literal service to TikTok
and hosting for TikTok,
0:33:19.697,0:33:22.432
you have probably
at this point accrued
0:33:22.433,0:33:27.537
at least half a trillion with a T,
half a trillion dollars of liability
0:33:27.538,0:33:32.809
that you cannot be sure
will not come due either
0:33:32.810,0:33:34.428
by Trump or a future president.
0:33:34.429,0:33:36.446
I will say, there's
a company called Akamai,
0:33:36.447,0:33:39.416
which is one of these internet
infrastructure companies
0:33:39.417,0:33:41.668
that also does business with TikTok.
0:33:41.669,0:33:45.539
They filed in their 10K
quarterly shareholder
0:33:45.540,0:33:48.408
report a really remarkable statement.
0:33:48.409,0:33:50.811
They said that they have accrued
substantial legal liability
0:33:50.812,0:33:54.631
for doing business with TikTok
and that they cannot be assured
0:33:54.632,0:33:57.368
that the Trump administration
or a future administration
0:33:57.369,0:34:03.057
will not impose
that liability on them.
0:34:03.174,0:34:05.358
You have a situation in which TikTok
0:34:05.359,0:34:10.581
is in this Schrödinger's
neither dead nor alive situation.
0:34:13.351,0:34:14.919
It's totally unclear
what's going to happen
0:34:14.920,0:34:19.456
because this is now wrapped up
in a massive trade war with China.
0:34:19.457,0:34:20.807
Of course, the Chinese government
0:34:20.808,0:34:24.978
is not going to give its permission
for ByteDance to fully divest
0:34:24.979,0:34:28.849
of TikTok while Trump is imposing
infinity plus one tariffs
0:34:28.850,0:34:31.953
or whatever it is as of this morning.
0:34:32.036,0:34:37.040
A bunch of giant American companies
have voluntarily exposed themselves
0:34:37.041,0:34:41.429
to potentially
company-ruinous liability
0:34:42.530,0:34:46.133
and are now subject
to profound blackmail,
0:34:46.134,0:34:47.551
for lack of a better term,
0:34:47.552,0:34:52.256
by this administration that has,
as with many other areas,
0:34:58.045,0:35:00.647
basically decided that it gets
to choose which laws to enforce.
0:35:00.648,0:35:04.134
What's amazing
is that this has totally overshadowed
0:35:04.135,0:35:06.787
the incredibly big
First Amendment stakes
0:35:06.788,0:35:07.855
of the actual decision.
0:35:07.856,0:35:10.891
No one is actually talking
about anymore
0:35:10.892,0:35:13.343
the fact
that the Supreme Court issued
0:35:13.344,0:35:17.632
an extremely important,
truly, for the case books,
0:35:17.698,0:35:21.186
opinion on the First
Amendment speeches
0:35:21.203,0:35:23.870
that applies to foreign platforms.
0:35:23.871,0:35:25.139
No one's talking about that
0:35:25.140,0:35:26.740
because it's been
totally overshadowed
0:35:26.741,0:35:30.377
by this rule of law story
that's now happening.
0:35:30.378,0:35:33.079
Sorry, I have many
feelings about TikTok
0:35:33.080,0:35:35.632
that I'm clearly
working out in real time.
0:35:35.633,0:35:36.750
My apologies to the audience
0:35:36.751,0:35:39.719
for this impromptu
therapy session for me.
0:35:39.720,0:35:42.365
Thanks for sharing, fascinating.
0:35:46.277,0:35:49.946
I'm definitely not a go-to expert
on the TikTok ban,
0:35:49.947,0:35:54.167
but I agree that the TikTok ban
is a source of political leverage.
0:35:54.168,0:35:57.437
It's an extremely important story
here, really troubling.
0:35:57.438,0:36:02.275
Also realizing I have to slightly
correct something I said earlier,
0:36:02.276,0:36:03.227
I attributed the back
0:36:03.228,0:36:05.178
to the Salt Mines tweets
to Dave Portney.
0:36:05.179,0:36:06.379
It was not him.
0:36:06.380,0:36:07.264
He was the subject
0:36:07.265,0:36:10.200
of a different NLRB enforcement
action arising out of the tweet.
0:36:10.201,0:36:12.986
It was actually Ben Domenech
who gave the salt mines tweet.
0:36:12.987,0:36:15.121
I don't want to wrongly attribute
0:36:15.122,0:36:17.207
the potentially labor
law-infringing tweets
0:36:17.208,0:36:19.876
to the wrong Twitter account here.
0:36:19.877,0:36:22.112
First Amendment lawyers
are particularly sensitive
0:36:22.113,0:36:29.113
to [laughs] accurate statements [?]
0:36:29.153,0:36:32.689
Since we talked a little bit
about international environment
0:36:32.690,0:36:36.443
with TikTok,
maybe now would be a good time
0:36:36.444,0:36:37.861
to talk a little bit about
0:36:37.862,0:36:42.749
the fact that US-based social media
platforms operate
0:36:42.750,0:36:45.802
in a global environment
where regulations from key markets
0:36:45.803,0:36:50.090
like the EU and UK
impact operational practices
0:36:50.091,0:36:52.909
and data handling
and content policies
0:36:52.910,0:36:56.084
and safety features
in those geographies.
0:36:56.814,0:36:59.849
Are international
social media regulations
0:36:59.850,0:37:03.755
inevitably impacting
US-based practices too?
0:37:04.122,0:37:07.742
Any thoughts or comments on that?
0:37:13.948,0:37:16.967
I think there is a lot
of talk about the GDPR,
0:37:16.968,0:37:19.786
and I certainly click I agree
0:37:19.787,0:37:22.889
more often
when I'm visiting a website.
0:37:22.890,0:37:24.057
That's exactly
what I was going to say.
0:37:24.058,0:37:26.372
There's a lot more clicking.
0:37:29.330,0:37:33.717
It's hard for me to say
that there's more effects than that.
0:37:33.718,0:37:39.506
Look, I think there's a lot
of potential for more conflict
0:37:39.507,0:37:44.311
between the US and the EU about
social media platform regulation.
0:37:44.312,0:37:48.783
Certainly, there's been
some signals about that.
0:37:49.583,0:37:54.020
The EU adopts
a more regulatory approach,
0:37:54.021,0:37:57.557
especially with some categories
of harmful speech,
0:37:57.558,0:38:00.573
such as some categories
of hate speech.
0:38:01.979,0:38:07.100
I think it's likely
that there will be some US government
0:38:07.101,0:38:10.021
more saber rattling about that.
0:38:10.888,0:38:15.569
I'm not sure I've got a prediction
about how that plays out.
0:38:15.893,0:38:17.444
One thing I will say
is that there's definitely
0:38:17.445,0:38:20.280
I think room
for social media companies
0:38:20.281,0:38:22.699
and American tech
companies to push back.
0:38:22.700,0:38:24.985
This is not a social media
example exactly,
0:38:24.986,0:38:27.270
but I think it is somewhat related.
0:38:27.271,0:38:32.994
OpenAI, for example, delayed
many of its offerings by weeks,
0:38:33.011,0:38:35.412
if not months in various EU countries
0:38:35.413,0:38:36.997
because they said,
0:38:36.998,0:38:39.566
"Look, we have this really cool
voice assistant,
0:38:39.567,0:38:45.121
but we can't deploy that in the EU
because of all sorts of regulations.
0:38:45.806,0:38:47.541
What they were not willing to do
0:38:47.542,0:38:53.547
was to cripple their US services
just to maintain parity with Europe.
0:38:53.814,0:38:59.570
I think there's this constant push
and pull between the Europeans who,
0:39:00.438,0:39:03.656
the charitable reading is care about
0:39:03.657,0:39:06.676
the data protection or data privacy
or whatever of their citizenry
0:39:06.677,0:39:09.029
and are not going
to compromise on that.
0:39:09.030,0:39:10.096
The less charitable reading
0:39:10.097,0:39:12.983
is can't produce
their own tech sector,
0:39:12.984,0:39:17.982
so they've decided to just
regulate other people's tech sectors.
0:39:18.873,0:39:20.940
Elements of both of
that are probably true.
0:39:20.941,0:39:22.142
There's this constant push and pull
0:39:22.143,0:39:26.396
between how much
of this are the US tech companies
0:39:26.397,0:39:28.048
going to swallow and how much
of this are going to say,
0:39:28.049,0:39:29.549
"No, we're just not
going to do business
0:39:29.550,0:39:33.837
in Europe on this
particular feature,"
0:39:33.838,0:39:36.317
and call the Europeans' bluff.
0:39:36.957,0:39:39.626
If a US company
turns off its services
0:39:39.627,0:39:42.479
in Europe, sure,
it's going to harm the US company,
0:39:42.480,0:39:46.015
but they're going to be a lot
of angry Europeans calling their--
0:39:46.016,0:39:47.867
I'm not exactly sure how it works,
their member
0:39:47.868,0:39:51.835
of the European parliament
or whatever the case is.
0:39:51.839,0:39:55.458
There's chicken being played
on both sides here, I think.
0:39:55.459,0:39:58.445
Obviously, the balance
between regulation
0:39:58.446,0:40:00.597
and stifling innovation.
0:40:00.598,0:40:03.817
which is the attention
that I think we all feel.
0:40:03.818,0:40:06.569
I was just traveling abroad,
and Charlotte, to your point,
0:40:06.570,0:40:08.571
found myself clicking, I agree,
0:40:08.572,0:40:12.258
thousands of times just
to get the most basic interaction
0:40:12.259,0:40:14.095
with technology there.
0:40:14.695,0:40:17.580
That's something
that I hope is not coming
0:40:17.581,0:40:18.948
to a theater near us here
0:40:18.949,0:40:22.599
in the United States,
but we'll see, I suppose.
0:40:24.388,0:40:27.358
Talked a little bit about challenges
0:40:28.008,0:40:30.908
and different regulatory approaches.
0:40:32.246,0:40:35.215
Maybe you could talk a little bit
more about the inherent challenges
0:40:35.216,0:40:39.203
in regulating a medium
as dynamic as social media.
0:40:39.420,0:40:43.039
I think many of us think
of it as decentralized,
0:40:43.040,0:40:46.309
but is it, or is it
really quite centralized,
0:40:46.310,0:40:50.356
and is that part
of the conversation or the problem?
0:40:50.514,0:40:53.149
I don't think
it's particularly decentralized.
0:40:53.150,0:40:56.386
There are certainly
decentralized alternatives.
0:40:56.387,0:40:58.138
I got very excited and wrote
a paper about
0:40:58.139,0:41:00.924
Mastodon when that was coming out,
0:41:00.925,0:41:02.276
because I thought that was an--
0:41:02.277,0:41:04.260
and it's very interesting,
0:41:04.261,0:41:05.529
and certainly
I think decentralization
0:41:05.530,0:41:07.130
is growing a little bit.
0:41:07.131,0:41:08.699
At the end of the day,
0:41:08.700,0:41:11.217
there are only four
or five companies.
0:41:11.218,0:41:13.780
The reason is that social media
0:41:16.073,0:41:17.992
naturally consolidates.
0:41:18.509,0:41:20.310
Some of this is because
of monopolistic practices,
0:41:20.311,0:41:22.813
but mostly it's actually
because of the dynamics
0:41:22.814,0:41:24.497
of how social media works.
0:41:24.498,0:41:27.750
Which is that social media platforms
have what are called network effects,
0:41:27.751,0:41:31.104
where the value of the platform
is not just about its features
0:41:31.105,0:41:32.088
or its content moderation,
0:41:32.089,0:41:34.858
but it's also about
other people on the platform.
0:41:34.859,0:41:38.611
What that naturally leads
to is winner-take-all effects,
0:41:38.612,0:41:42.900
where naturally, once a platform
0:41:42.901,0:41:44.851
is slightly
better than its competitors,
0:41:44.852,0:41:46.135
it naturally rises to the top,
0:41:46.136,0:41:49.310
and so you have a power
law distribution.
0:41:49.790,0:41:52.075
Where you have this very small number
0:41:52.076,0:41:53.543
of giant platforms
clustered at the top,
0:41:53.544,0:41:56.729
and you have a very long tail
of cool niche platforms
0:41:56.730,0:42:01.134
that don't really matter
that much at the end of the day.
0:42:01.135,0:42:07.074
Centralization makes it both easier
and harder to regulate.
0:42:07.191,0:42:08.341
It makes it easier to regulate
0:42:08.342,0:42:11.833
in that you're only
regulating a few targets.
0:42:14.298,0:42:16.216
If we were to fund this adequately,
0:42:16.217,0:42:18.952
and everyone constantly complains
that the government's too slow.
0:42:18.953,0:42:19.969
Most of the government's too slow
0:42:19.970,0:42:21.321
because we just don't fund
it sufficiently,
0:42:21.322,0:42:22.622
or the government's too slow
because we've decided
0:42:22.623,0:42:23.907
to make the government too slow,
0:42:23.908,0:42:28.111
and the antics of the last couple
of months are not helping with that.
0:42:28.112,0:42:30.496
Yes, right now, the government
can't regulate anything
0:42:30.497,0:42:33.417
at the moment, but that's very much
a choice that we have made,
0:42:33.418,0:42:36.703
and maybe we will unmake
it in two or four years time.
0:42:36.704,0:42:38.622
On the one hand,
the centralized nature
0:42:38.623,0:42:40.440
of social media makes
it easier to regulate.
0:42:40.441,0:42:44.510
On the other hand, it also means
that you have a small group
0:42:44.511,0:42:48.782
of players
who are very well-resourced.
0:42:48.933,0:42:50.633
These are literally the biggest
companies in the world.
0:42:50.634,0:42:52.352
They're very incentivized
to collaborate,
0:42:52.353,0:42:56.189
to push back against regulation
0:42:56.190,0:42:57.557
because you have a classic situation
0:42:57.558,0:43:01.194
where the benefits
of regulation are diffuseed
0:43:01.195,0:43:04.448
to the extent you think
that they're good,
0:43:04.648,0:43:07.283
they come to all of us,
as individuals,
0:43:07.284,0:43:09.235
and so none of us
individually has much
0:43:09.236,0:43:11.070
of an incentive
to really lobby Congress.
0:43:11.071,0:43:14.574
The cost of that regulation
fall on a very small number
0:43:14.575,0:43:17.344
of regulated parties,
and so they have the incentive
0:43:17.345,0:43:19.078
and the coordination
ability to push back.
0:43:19.079,0:43:20.446
Now, if I were a regulator,
0:43:20.447,0:43:26.152
I would still much rather
have a smaller field.
0:43:26.153,0:43:29.389
By smaller, I mean a field
with fewer participants
0:43:29.390,0:43:31.208
to get my arms around.
0:43:31.342,0:43:32.492
It does pose challenges
0:43:32.493,0:43:37.080
because you're some
mid-level civil servant
0:43:37.081,0:43:41.117
in the Federal Trade Commission,
and your counterparty is Meta,
0:43:41.118,0:43:43.620
which has infinite,
last time I checked,
0:43:43.621,0:43:47.588
I think literally infinite
dollars at its disposal.
0:43:49.943,0:43:53.746
This is all with the First
Amendment hanging over.
0:43:53.747,0:43:56.849
Again, I cannot emphasize
enough how uncertain,
0:43:56.850,0:43:58.418
forget the regulatory landscape is,
0:43:58.419,0:44:02.389
just the basic
constitutional landscape is.
0:44:02.756,0:44:07.343
I think if you're a regulator
right now, you're in the same level
0:44:07.344,0:44:10.380
of constitutional
uncertainty as if you're FDR
0:44:10.381,0:44:12.632
in 1938 with respect
to the Commerce Clause.
0:44:12.633,0:44:14.817
It is that level of, I have no idea
0:44:14.818,0:44:16.436
what I am and am not allowed to do,
0:44:16.437,0:44:19.672
and the only way I can figure
it out is just by regulating
0:44:19.673,0:44:21.174
and then litigating for five years.
0:44:21.175,0:44:23.059
Which is the way
it's supposed to work,
0:44:23.060,0:44:25.412
but makes it very difficult.
0:44:25.763,0:44:27.481
It's great for people
like Charlotte and me
0:44:27.482,0:44:29.582
who make our lives writing
larvae articles about this.
0:44:29.583,0:44:33.153
It's not so great if you're
an FTC bureaucrat.
0:44:35.789,0:44:39.108
Is it different if you're
a state-level regulator?
0:44:39.109,0:44:41.928
You mentioned
the content moderation laws.
0:44:41.929,0:44:45.499
Is that likely to be an area
of more activity?
0:44:48.235,0:44:52.939
Yes, and similar to what
Alan was saying before,
0:44:52.940,0:44:54.240
especially with respect to children.
0:44:54.241,0:44:58.344
A bunch of states
have recent statutes
0:44:58.345,0:45:02.783
that affect how children
use social media.
0:45:03.350,0:45:08.688
Things like,
if a minor is using social media,
0:45:08.689,0:45:10.724
the platform has to limit
0:45:10.725,0:45:12.931
how they communicate
with other people,
0:45:12.932,0:45:13.994
especially other people
0:45:13.995,0:45:16.643
who they aren't directly
friends with,
0:45:16.644,0:45:19.288
or aren't directly connected to.
0:45:19.883,0:45:22.018
There's a lot of litigation
0:45:22.019,0:45:25.272
at the preliminary
injunction stage there.
0:45:25.823,0:45:28.441
I think the proliferation
of those statutes,
0:45:28.442,0:45:30.710
and I agree
their future is uncertain,
0:45:30.711,0:45:34.797
is a signal that when states
feel really motivated
0:45:34.798,0:45:36.466
to do something about something
0:45:36.467,0:45:39.018
they perceive
as a social media-related problem,
0:45:39.019,0:45:41.671
they can move reasonably quickly.
0:45:41.672,0:45:44.207
Whether they move carefully enough,
0:45:44.208,0:45:48.254
to whether the things
they're doing are a good idea,
0:45:48.896,0:45:50.576
is another question.
0:45:51.231,0:45:53.099
They can definitely move things out
0:45:53.100,0:45:56.353
of their legislatures
and into the courts.
0:45:58.338,0:46:03.292
Likely leading
to challenges in state courts
0:46:03.293,0:46:07.831
that ultimately might rise
to the federal court as well.
0:46:08.298,0:46:11.033
Is that where you see
a lot more activity
0:46:11.034,0:46:15.355
similar to the NetChoice cases,
that process?
0:46:16.207,0:46:17.207
Yes.
0:46:17.875,0:46:21.260
Getting ready for this,
I think I've realized
0:46:21.261,0:46:22.879
that you can find all these cases
0:46:22.880,0:46:26.549
by going into Westlaw and searching
by party name for NetChoice.
0:46:26.550,0:46:29.519
That quickly brings up this long list
0:46:29.520,0:46:33.089
of mostly preliminary
injunction cases,
0:46:33.090,0:46:34.824
a lot of them in federal court,
0:46:34.825,0:46:38.394
raising First Amendment questions
about these attempts
0:46:38.395,0:46:41.280
to regulate either content moderation
0:46:41.281,0:46:45.327
in general or specifically
with respect to children.
0:46:47.688,0:46:51.417
I don't think
we have any questions in the chat.
0:46:51.742,0:46:54.469
Actually, we have one in the Q&A.
0:46:56.630,0:47:00.280
I am not seeing that,
so if you could go ahead.
0:47:00.601,0:47:02.281
Okay, here we go.
0:47:02.719,0:47:05.371
The question in the Q&A,
can we talk about government entities
0:47:05.372,0:47:06.506
that are on social media
0:47:06.507,0:47:09.091
and what the law says
about content moderation
0:47:09.092,0:47:11.627
on their social media
pages and posts?
0:47:11.628,0:47:14.363
Can a state, federal,
or local government entity choose
0:47:14.364,0:47:18.384
to allow comments on some,
but not all of its posts?
0:47:18.385,0:47:21.654
Are there any restrictions
on a government entity choosing
0:47:21.655,0:47:24.540
to cut off comments
on their social media pages
0:47:24.541,0:47:29.857
and posts altogether if the comments
are too many or becoming toxic?
0:47:30.197,0:47:31.699
In those--
0:47:31.716,0:47:33.083
Do you want to jump
on this one, Charlotte?
0:47:33.084,0:47:35.184
Yes, in those cases,
the question is often
0:47:35.185,0:47:37.737
whether the government
has created a forum.
0:47:37.738,0:47:42.542
If you remember forum analysis
from general First Amendment law,
0:47:42.543,0:47:45.270
you can apply that question here.
0:47:45.812,0:47:49.550
When government
creates a limited forum,
0:47:51.051,0:47:52.585
its restrictions
have to be reasonable
0:47:52.586,0:47:54.405
and viewpoint-neutral.
0:47:54.988,0:47:58.224
When government creates a forum,
it can also cut off the forum.
0:47:58.225,0:48:01.911
If government wants to use
social media as a one-way vector
0:48:01.912,0:48:03.946
for information to go out,
but doesn't want
0:48:03.947,0:48:08.401
to receive commentary back,
that's, I think, generally fine.
0:48:08.402,0:48:10.620
Where government entities
0:48:10.621,0:48:14.691
and where individual office
holders run into trouble
0:48:14.958,0:48:18.813
is when they either selectively
0:48:19.413,0:48:24.750
delete comments or selectively delete
0:48:24.751,0:48:27.053
or selectively accept
comments based on viewpoint,
0:48:27.054,0:48:31.040
or in an unreasonable way, or say,
0:48:31.041,0:48:33.743
"This individual
constituent is blocked
0:48:33.744,0:48:35.778
and therefore
they can't access this source
0:48:35.779,0:48:37.280
of government information."
0:48:37.281,0:48:39.348
There can be a preliminary
question in those cases about
0:48:39.349,0:48:42.418
whether the account is really
a governmental account or not.
0:48:42.419,0:48:48.441
If it is, then I would say generally
the government's approach
0:48:48.442,0:48:51.027
to handling that account
has to be viewpoint-neutral
0:48:51.028,0:48:53.094
and has to be reasonable.
0:48:53.747,0:48:55.581
I would just add,
there was a case about
0:48:55.582,0:48:59.352
that preliminary question last year
0:48:59.353,0:49:04.524
in a case called Lindke v. Freed.
0:49:05.042,0:49:07.811
This was asking the question
0:49:07.812,0:49:10.030
of when is a government
official's use
0:49:10.031,0:49:13.666
of social media state action,
such as you even ask
0:49:13.667,0:49:17.837
all the questions Charlotte just
raised about limited public forum
0:49:17.838,0:49:20.606
and viewpoint neutrality
and stuff like that.
0:49:20.607,0:49:24.477
This comes up because often,
you're seeing
0:49:24.478,0:49:26.957
this now at the highest levels
0:49:27.331,0:49:31.139
with Donald Trump on
Truth Social and J.D. Vance,
0:49:31.152,0:49:32.602
who posts a lot.
0:49:32.603,0:49:34.854
A lot of government
officials are using social media
0:49:34.855,0:49:36.222
and sometimes it's not clear
0:49:36.223,0:49:40.042
whether they're using
it as government officials,
0:49:40.043,0:49:41.844
whether they have
the authority to do so.
0:49:41.845,0:49:43.313
Obviously, with Trump and Vance,
0:49:43.314,0:49:45.031
it's obvious that they have
the authority to do so,
0:49:45.032,0:49:47.717
but alternatively,
0:49:47.718,0:49:49.969
whether they're using
it just in their personal capacities
0:49:49.970,0:49:52.036
and whether it's blended.
0:49:53.090,0:49:54.957
What the court says, you have to look
0:49:54.958,0:49:59.930
on a per host basis
about what specifically--
0:50:00.380,0:50:01.998
was the government official speaking
0:50:01.999,0:50:04.500
as a government official
or as an individual?
0:50:04.501,0:50:05.751
If they're speaking
as a government official,
0:50:05.752,0:50:10.790
then yes, the reply
thread under that post
0:50:10.791,0:50:13.943
is potentially a limited
public forum,
0:50:13.944,0:50:16.679
and then you do all this stuff
Charlotte was talking about.
0:50:16.680,0:50:18.864
If it was just,
"Hey, I had tacos for breakfast,"
0:50:18.865,0:50:20.716
and the response is, "You stink,"
0:50:20.717,0:50:23.819
then there's an argument that no,
0:50:23.820,0:50:26.455
he was just speaking
in a personal capacity.
0:50:26.456,0:50:29.208
I think what's tricky
is that both is very fact specific,
0:50:29.209,0:50:33.613
and also, Lindke v. Freed,
the Supreme Court case was about--
0:50:33.614,0:50:35.281
I looked this up just now,
0:50:35.282,0:50:36.415
it was about the city manager
0:50:36.416,0:50:38.901
of Port Huron, Michigan,
which is to say,
0:50:38.902,0:50:41.420
not a high-level government official.
0:50:41.421,0:50:42.789
There, I think it's
more reasonable to say,
0:50:42.790,0:50:43.739
"Look, he's a government official,
0:50:43.740,0:50:44.774
but he's also just some dude,
0:50:44.775,0:50:47.927
and so he is allowed
to also have a personal life."
0:50:47.928,0:50:49.395
I'm not sure
the President of the United States
0:50:49.396,0:50:50.696
is allowed to have a personal life,
0:50:50.697,0:50:51.864
except for being president
of the United States.
0:50:51.865,0:50:54.051
That's the whole point.
0:50:55.052,0:50:58.187
The rules for when something
becomes government speech
0:50:58.188,0:51:01.140
and state action might be different,
0:51:01.141,0:51:03.175
depending on how high up
the food chain
0:51:03.176,0:51:07.681
of government officialdom
you're looking at.
0:51:09.800,0:51:11.784
Another question that I'll tee up
0:51:11.785,0:51:15.171
for you both that came up in the Q&A.
0:51:15.172,0:51:17.389
Are there any First Amendment
implications attached
0:51:17.390,0:51:19.341
to how states and federal
government regulate data
0:51:19.342,0:51:22.378
that social media
companies collect from users,
0:51:22.379,0:51:25.548
like the California
Consumer Privacy Act?
0:51:25.549,0:51:28.617
Any First Amendment implications
attached to requiring companies
0:51:28.618,0:51:32.906
to delete data
when a consumer requests it?
0:51:35.542,0:51:37.877
I think there's a simple answer,
which is, yes,
0:51:37.878,0:51:39.411
there are First Amendment
implications to this.
0:51:39.412,0:51:41.180
What the First Amendment
implications are,
0:51:41.181,0:51:43.247
that's the hard question.
0:51:44.251,0:51:46.702
We obviously have a ton
of privacy statutes
0:51:46.703,0:51:47.903
generally in the United States,
0:51:47.904,0:51:50.089
both in the federal
and the state level.
0:51:50.090,0:51:53.692
Those generally have not been thought
necessarily even to implicate
0:51:53.693,0:51:55.728
the First Amendment,
let alone violate it.
0:51:55.729,0:51:57.446
Everyone's heard of HIPAA.
0:51:57.447,0:52:00.032
No one thinks that HIPAA
is a First Amendment issue.
0:52:00.033,0:52:00.616
It is what it is.
0:52:00.617,0:52:03.235
You can argue pro-con,
but no one's like,
0:52:03.236,0:52:06.722
"It's a First Amendment problem
that the government restricts
0:52:06.723,0:52:09.091
the transmission
of health information."
0:52:09.092,0:52:12.678
Charlotte and I work at a university,
so we're limited very much
0:52:12.679,0:52:16.549
of what we can say about
our students under, I think, FERPA,
0:52:16.550,0:52:20.187
one of these student
data privacy statutes.
0:52:20.670,0:52:24.740
Then, in the 2000s or 2010s,
I forget when, the court threw
0:52:24.741,0:52:28.460
a bomb into all of this
with this opinion called Sorrell.
0:52:28.461,0:52:32.532
That was about
the restriction on the sale
0:52:32.533,0:52:37.938
of health care
prescription prescriber data.
0:52:38.972,0:52:41.123
Depending on how broadly
or narrowly you read
0:52:41.124,0:52:44.660
that opinion,
either the data privacy regime
0:52:44.661,0:52:47.396
that has existed
is largely unchanged,
0:52:47.397,0:52:49.915
does not really implicate
First Amendment,
0:52:49.916,0:52:52.368
or every single data privacy regime
0:52:52.369,0:52:54.987
is now squarely
a First Amendment question
0:52:54.988,0:52:58.224
because it's about data,
and data is speech,
0:52:58.225,0:53:00.743
and now you're off to the races.
0:53:00.744,0:53:05.214
My unfortunate cop-out answer is yes,
but stay tuned.
0:53:05.215,0:53:07.466
This is one of the very long list
0:53:07.467,0:53:08.817
of digital First Amendment questions
0:53:08.818,0:53:14.273
that the courts will have to address
in the next decade or so.
0:53:14.274,0:53:17.076
I'm curious, Charlotte,
if you agree with how I framed that.
0:53:17.077,0:53:18.077
I do.
0:53:18.778,0:53:23.618
In the First Circuit, the case
that eventually became Sorrell,
0:53:23.717,0:53:25.351
the court went the other way.
0:53:25.352,0:53:27.870
One thing the court said in that case
0:53:27.871,0:53:30.206
is a regulation on the sale of data
0:53:30.207,0:53:34.043
is the same as a regulation
on the sale of beef jerky.
0:53:34.044,0:53:37.914
I have to say,
I like the beef jerky analysis.
0:53:38.398,0:53:40.966
I think I'm alone in the legal
academy in thinking
0:53:40.967,0:53:44.086
that that would be a reasonable
way to think about this case.
0:53:44.087,0:53:48.992
The Supreme Court said,
"No, the sale of data is speech.
0:53:49.409,0:53:51.577
This incidentally
is one of the opinions
0:53:51.578,0:53:53.512
where you start to get
the more liberal justices
0:53:53.513,0:53:57.266
in dissent saying
our deregulatory approach
0:53:57.267,0:54:01.869
to the First Amendment
is really leading in bad directions.
0:54:02.722,0:54:07.944
Nonetheless, the Sorrell court
opinion by Justice Kennedy,
0:54:08.778,0:54:13.399
a characteristically
hard-to-pin-down opinion,
0:54:13.400,0:54:15.836
I would say, suggests that
0:54:16.297,0:54:21.173
there is some relatively
broad protection
0:54:21.174,0:54:24.824
under the First Amendment
for the sale of data.
0:54:25.045,0:54:26.646
To the extent people
have questions about
0:54:26.647,0:54:30.316
what does this mean for,
for example, net neutrality?
0:54:30.317,0:54:32.714
That's an excellent question.
0:54:37.374,0:54:42.111
I think that's all the questions
that we have in the chat and the Q&A.
0:54:42.112,0:54:46.079
Any final comments or thoughts,
Alan and Charlotte?
0:54:49.069,0:54:51.654
I will just sort of say
what I came back to,
0:54:51.655,0:54:57.193
which is, as frustrating
as I find this area of law,
0:54:57.194,0:55:02.298
because so much is uncertain
and unclear, and so much
0:55:02.299,0:55:04.800
of the stuff that like I want
to do in AI regulation,
0:55:04.801,0:55:06.151
there's this giant,
0:55:06.152,0:55:09.538
but the First Amendment may be cloud
hanging over all that.
0:55:09.539,0:55:13.025
I actually find it also among
0:55:13.026,0:55:16.795
the more intellectually
exciting parts of the law.
0:55:16.796,0:55:19.198
Again, not just because so much
of it is a greenfield,
0:55:19.199,0:55:22.501
but because it does not match
0:55:22.502,0:55:25.914
neatly to political
or partisan preferences.
0:55:26.306,0:55:31.304
There's real thinking, real first
principles thinking happening.
0:55:31.678,0:55:33.980
That's always fun to see.
0:55:36.199,0:55:38.667
As an American, it gives
me a very warm, fuzzy feeling
0:55:38.668,0:55:45.408
when Kagan and Kavanaugh are yelling
at Alito and Sotomayor.
0:55:46.226,0:55:50.479
That's the kind of intra-court
fighting that I want to see.
0:55:50.480,0:55:55.769
That makes me feel
like our institutions are doing okay
0:55:56.336,0:55:58.070
because there's so much
other bad news
0:55:58.071,0:56:00.773
about our institutions, seeing
those kinds of fights rather than,
0:56:00.774,0:56:02.958
"Oh, you got the liberals
and the conservatives,"
0:56:02.959,0:56:04.043
that makes me feel good.
0:56:04.044,0:56:05.094
Despite my frustration,
0:56:05.095,0:56:07.730
I actually quite enjoy
playing in this space
0:56:07.731,0:56:11.200
and thinking about
all these legal questions.
0:56:11.201,0:56:15.020
I guess I'll say,
I agree with Alan's premises.
0:56:15.021,0:56:17.156
I read them slightly differently.
0:56:17.157,0:56:21.443
I guess I think
the partisan implications
0:56:21.444,0:56:24.538
in this area are sorting
themselves out.
0:56:25.498,0:56:28.952
I tend to think
that there is still a way
0:56:28.969,0:56:34.891
that partisanship is very important
in driving these cases.
0:56:36.593,0:56:40.964
I think I tend to see
more partisanship, for example,
0:56:43.550,0:56:45.384
in the various NetChoice opinions
0:56:45.385,0:56:49.422
than maybe Alan does.
0:56:50.741,0:56:51.959
Who knows?
0:56:53.777,0:56:55.344
Thank you both for your insights.
0:56:55.345,0:56:56.812
Thanks, everyone, for attending.
0:56:56.813,0:57:00.799
I hope everyone enjoys
the rest of spring alumni week.
0:57:00.800,0:57:03.197
Have a great week, everybody.