0:03

Hello! Friends. I'm Jeffrey Rosen, president

0:05

Ceo The National Constitution Centre and

0:08

welcome the We The People, a

0:10

show of constitutional debate, the National

0:12

Constitution Centers and nonpartisan nonprofits charted

0:15

by Congress to increase awareness and

0:17

understanding of the Constitution among the

0:20

American people. In Twenty nineteen, Facebook

0:22

surpass two billion users, and the

0:24

company created an independent oversight board

0:27

known as the Supreme Court a

0:29

Facebook to review appeals of it's

0:31

content decisions. In this episode, Members

0:34

of the Supreme Court a Facebook

0:36

which is now known as The

0:38

Metal Oversight Board, join me to

0:40

discuss the board structure, it's key

0:42

decisions, and it's efforts to ensure

0:44

free and fair elections. In Advance

0:46

of Twenty Twenty Four, it was

0:48

an honor to convene. Michael Mcconnell

0:51

of Stanford Law School and Kenji

0:53

Yoshino of New York University School

0:55

of Law. Enjoyed the conversation. It

0:58

is a great pleasure and an

1:01

honor for the Ansi seat to

1:03

convene to great scholars to discuss

1:05

the role of Mehta's oversight board.

1:08

And questions involving

1:10

election integrity. ah.

1:13

Michael. Mcconnell is Richard and

1:15

Francis Mallory professor and director

1:17

of the Constitutional Law Center

1:19

at Stanford Law School and

1:22

a senior fellow at the

1:24

Hoover Institution. Kenya Sheena, his

1:26

Chief Justice or Warren Professor

1:28

Constitutional Law and Why You

1:30

and Director of the Meltzer

1:32

Center for Diversity, Inclusion and

1:34

Belonging are a good. Professors

1:37

Mcconnell and Yoshino are.to America's

1:39

leading constitutional scholars. They come

1:41

from diverse perspectives, but there.

1:44

Are united in a commitment to

1:46

the First Amendment and in their

1:48

service on the matter. Oversight Board

1:50

which is a unique body with

1:52

a with a great responsibility for

1:55

enforcing free speech values in times

1:57

of elections. I was so honored

1:59

when. When they agreed to come

2:01

to tell us about what's the

2:04

matter, Oversight Board does and to

2:06

discuss some of it's important. Recent

2:08

cases involving election integrity and I

2:10

know we all have a lot

2:12

or to learn from them are

2:15

welcome. I'm a keynesian. Michael are

2:17

kenji of what did you start

2:19

by telling us what? the Mehta

2:21

Oversight Board. Does and what

2:23

sort of issues involving elections it

2:25

deals with. Yeah so first

2:28

of all of it's an enormous pleasure to be

2:30

with you Here today at of from privilege and

2:32

on our our is mine. So. The

2:34

Matter Oversight Board as a body

2:37

and funny to individuals globally of

2:39

who helped Madoff deal with the

2:41

thorniest contact. Mater is an issue

2:43

thought that Platform faces. Were.

2:45

Are independent body or that stands

2:48

apart from the corporation itself. Or

2:50

and we make our decisions and

2:53

seven priority areas. Elections

2:55

and civil Space crisis and

2:57

complex situation. Sundar. Hate.

2:59

Speech Government use about

3:01

of platforms Automation. And.

3:04

Pretty users fairly. So. I

3:06

think we're coming together today to talk about

3:08

that. A lot Fans Priority. And. Under

3:10

that back at we have a weather

3:13

eye out for a number of issues

3:15

but prominent among those are things like

3:17

of missing from a some or during

3:19

election. Other: the process of

3:21

dissidents or political voices. I

3:23

would be another big one. Or and

3:26

then finally violence and inside matt near the

3:28

worry that in out these elections which to

3:30

be safe and. Ah, As

3:32

a car around the world are gonna be

3:34

subjected to threats of violence are actual violence?

3:37

The thought. Of the remiss if

3:39

I didn't conclude by saying bad enough as

3:41

we all now this is our ultimate Alex

3:43

and yet or forty nine percent of the

3:45

global population of going to the polls. About

3:48

I do not think that we have ever

3:50

seen on Alex in your life of before

3:52

and fell in that context or the work

3:54

becomes all the more important. Doesn't

3:57

date and were very glad to have the chance to.

4:00

discuss it. Michael, the the

4:02

Oversight Board was a unique structure when

4:04

it was created. You

4:06

were present from the

4:08

creation. Tell us what the

4:10

goals were in creating the Oversight Board

4:12

was and how it's worked. Well

4:16

Jeff, the idea here is that

4:19

the social media

4:21

companies are enormously important

4:24

and powerful gatekeepers for

4:27

speech. And some people

4:29

think well why don't they just let everything go

4:31

but nobody actually wants that. We all want spam

4:34

out. We all want the law to

4:36

be enforced.

4:41

We don't want terrorists to be able to

4:43

organize over the network. We don't want people

4:45

to be bullied and

4:47

harassed. There are lots of

4:49

limitations on speech that are,

4:51

I think most users agree

4:53

about, but but having a

4:55

private company with its profit

4:59

motives and its interest in pressure

5:01

from advertisers, being in complete

5:04

control, strikes many people in

5:06

a democratic system as being

5:09

troublesome. And yet the government

5:11

as an alternative could easily

5:13

be worse. The

5:16

experience of government regulation of speech

5:18

around the world is not a

5:20

happy one. And so this is

5:24

an idea, an experiment, to

5:26

ask a certain number of

5:28

people on a part-time basis,

5:30

people who all have been

5:33

engaged in civic action

5:35

and various roles to

5:37

be an independent board.

5:39

We don't work for meta. None

5:42

of us has ever worked for

5:44

meta. Many people actually have

5:46

been critics of meta in the past and

5:50

to review specific cases and make

5:52

recommendations for how a meta can

5:55

deal with some of

5:57

its really hard problems. So

6:00

let's discuss some of the recent

6:02

decisions relating to free speech and

6:05

election integrity. Our friends

6:07

in the audience can find

6:09

the decisions on the website

6:12

of oversightboard.com. We'll chat some

6:15

of them. And

6:17

among the most important

6:19

initial ones involved

6:22

the decision to

6:24

uphold Betta's decision temporarily

6:27

to remove President Trump for

6:29

his comments surrounding

6:31

January 6th, but to reverse the

6:33

decision to suspend him indefinitely and

6:36

to insist on the adoption of

6:38

standards for deplatforming

6:41

politicians. Kenji

6:43

tell us about this decision, a kind of

6:45

Marbury versus Madison for the meta oversight board.

6:48

What did it hold and what is its

6:50

significance? Yes, absolutely. So

6:53

this is a violence and incitement

6:55

case where President Trump was deemed

6:58

by Metta to have incited violence with

7:00

his comments around the January 6th incident

7:03

at the Capitol. And

7:06

as you said, they suspended

7:09

his account indefinitely

7:12

without any kind of guideline

7:14

around that. So

7:16

I think oftentimes the headline

7:18

of what the oversight board

7:20

in that decision is not

7:23

really the most important thing that

7:25

we did. So the headline

7:27

was that we upheld Metta's decision to

7:30

suspend. I actually think that

7:32

the most important thing that we did in that case

7:34

was to say it is not

7:36

OK for Metta to create

7:39

an indefinite suspension. There have to

7:41

be some guidelines so that people

7:43

are aware of what

7:45

they have done and what the penalties are

7:47

and when they can expect the penalties to

7:51

be lifted. So as

7:53

Michael was saying, what we worry about

7:55

as the oversight board is that the

7:58

speech platform that has been of

8:01

users on it is being regulated

8:03

too much by economics and

8:05

profit motives rather than through international

8:07

human rights standards. So in

8:10

our decision making we sort of

8:12

adhere to certain international human rights

8:14

standards, particularly those promulgated under the

8:17

International Covenant of Civil and Political Rights

8:19

Article 19, which thinks

8:21

about things like legality, legitimate

8:23

aim, and necessity and proportionality.

8:26

And here the idea of legality is,

8:28

you know, issues that will be very

8:30

very familiar to you and many of

8:32

our listeners today through the US

8:35

Due Process Clause, which is that you have

8:37

kind of noticed an opportunity to be heard,

8:39

right, and you have a right to understand

8:41

the kind of penalties that have been leveled

8:43

against you. Also ideas like no kind of

8:46

ex post facto laws of saying we have

8:48

a right to notice, right, we have a

8:50

right not have vague standards

8:52

be applied to us. And so it's really

8:55

those kinds of impulses that refracted not

8:57

through our US constitutional framework because it's

8:59

the global standard, but rather through the

9:01

ICCPR Article 19 standard

9:03

of legality that pushed us

9:06

to say this has

9:09

to be a limited suspension to two years and

9:11

then you have to be clear going forward both

9:13

about, you know, this particular individual but with regard

9:15

to all users as to why penalties are being

9:17

leveled against them and when they can be expected

9:20

to be lifted. Very

9:22

interesting and important.

9:26

Michael, say more about

9:28

the speech that the board

9:30

held could be temporarily restricted, in particular

9:32

President Trump's comments, we love you, you're

9:34

very special in his first post, and

9:37

great patriots remember this day forever

9:39

in the second post violated Facebook's

9:41

rules prohibiting praise or support of

9:43

people engaged in violence,

9:47

but it wasn't permissible to impose an

9:49

indefinite suspension. There

9:51

was a minority

9:54

of the board that emphasized that Facebook

9:56

should take steps to prevent the repetition

9:58

of adverse human rights impacts and ensure

10:00

that users who seek reinstatement after suspension

10:02

recognize their wrongdoing and commit to observing

10:04

the rules in the future. The fact

10:06

that there's kind of a dissenting or

10:08

concurring opinion makes me ask, how

10:11

do the deliberations work? Do you meet by

10:13

Zoom or in person? Do you

10:15

talk by email and tell us

10:17

more about the significance of the

10:19

Trump decision? Well, first, how do

10:21

we work? We're

10:23

all over the world. There are 22 members,

10:26

every continent, except for Australia.

10:29

I'm sorry, except for Antarctica. We

10:31

do have someone from Australia. Time

10:35

zones are a real problem. It's

10:38

not infrequent for one person to be on the

10:40

west coast of the United States at 6 a.m.

10:43

and somebody else is often at

10:46

the eastern edge of Eurasia,

10:48

you know, very late

10:50

at night approaching midnight.

10:52

It's difficult. We do

10:54

meet together about twice a

10:56

year. That's

10:59

usually more organizational and sort

11:01

of to work things out

11:03

rather than to deliberate over

11:05

particular matters. We

11:08

have a practice of appointing a

11:10

five-person panel. There will

11:13

be one member. It's mostly random,

11:16

but with a guarantee of at

11:19

least one member from the region

11:22

most affected by the decision.

11:24

And then their decision

11:26

is circulated to the

11:29

entire 22-member group, and

11:31

there are comments and sometimes disagreements

11:35

there. And so I

11:38

actually think that the

11:40

practice of dissenting is very

11:42

important because it helps people

11:44

understand that we are actually

11:46

thinking through. And

11:49

these are not easy questions. And there

11:52

are no ready answers to them. And

11:55

I think by telling the

11:57

world what those things are, it's very important to understand

12:00

more serious disagreements are

12:02

that we actually advance

12:04

the cause. Now

12:07

you asked for some more detail

12:09

on the specifics. You quoted some

12:11

of the specific meta,

12:14

you know, meta meaning Facebook

12:17

and Instagram and now threads,

12:19

so there are three platforms

12:21

covered by that term, that

12:24

the community standards, the

12:27

specific ones that were most in

12:30

play with President Trump

12:33

on January 6th had to do

12:35

with his words of praise as

12:39

the riot was taking

12:41

place at the Capitol and members

12:43

of Congress were having to take

12:45

cover and, you know, the office

12:47

of the Speaker of the House

12:50

was invaded and people in

12:52

weird costumes were making threatening

12:56

statements. As

12:59

that's going on, the president

13:02

was actually praising them and

13:04

talking about how them

13:06

as patriots and so forth. And

13:09

so the community standard

13:11

on that came directly to bear.

13:15

That standard is also quite

13:17

important in connection with dangerous

13:19

organizations across the spectrum. So

13:21

this is not an anti-Trump

13:23

or, you know, specifically Trump

13:26

thing. The same issues come

13:28

into play if someone is praising Hamas

13:31

for what took place recently

13:34

in Israel or praising

13:37

the crushing of dissent in

13:40

Russia. Or it's

13:44

because one of the important ways in

13:47

which support is raised for,

13:49

you know, violence

13:51

and suppression of speech

13:54

is precisely through lauding

13:56

the actions of people engaged in

13:59

it. That was the

14:01

specific community standard

14:04

most important in the Trump decision. Thank

14:07

you for that. And that calling attention to that

14:10

community standard and its relevance around the

14:12

world is crucial. That

14:16

leads to a recent

14:18

decision overturning the

14:21

speech of a

14:23

Brazilian general. And

14:26

there, well, Kenji, why don't I

14:28

let you see that one

14:31

up and describe what the board held. Yeah,

14:33

so there's a hotly contested election

14:35

in Brazil. And I believe we

14:37

handed down this decision about a

14:40

year ago. And in the wake of that

14:43

decision election where

14:46

Bolsonaro lost,

14:49

there was a Brazilian general who in

14:52

uniform made a speech affording

14:54

individuals to essentially storm

14:56

the Capitol. So one of the interesting

14:58

things about being on this board is

15:01

that you see geopolitically how much particular

15:03

situations would rhyme with each other.

15:07

So there we again

15:10

took down the post

15:12

and said that you were not allowed to urge

15:15

individuals to engage in this

15:17

kind of activity

15:19

in a highly sensitive time,

15:22

in a highly sensitive area. And

15:25

so one of the things that we were

15:27

flagging there was a high sensitivity of election

15:30

issues. And we found that Matt had

15:32

done an insufficient job of protecting individuals

15:35

during election period.

15:38

And we urged them to engage in

15:40

greater sensitivity. So in some sense, what's

15:42

past this prologue, you know, that we were hopefully

15:45

pressured, right, and saying, we're going to

15:47

confront this over and over again. So

15:49

Matt needs to have teams

15:52

or attention played to particular

15:54

areas, particular time periods surrounding

15:56

elections so that these hotspots

15:58

that we know. will occur

16:00

with regard to speech, with regard

16:03

to, you know, incitement are more

16:05

within the control of the platform.

16:08

Can I just add to that that a

16:11

meta has what are called

16:13

crisis of policy

16:16

protocols in place when they expect

16:18

that there are going to be

16:20

this kind of, you

16:23

know, violence and commotion and

16:25

elections are treated under that

16:27

heading. But the

16:31

Brazilian case wasn't actually about

16:34

elections. It was about the post-election

16:36

period. And one of the things

16:38

we've stressed is that in these

16:40

days, the post-election period is

16:43

just as much prone to

16:45

this kind of action

16:49

as anything else, and that they need to

16:51

be aware of that. That

16:53

distinction between post-election and the

16:55

election itself is crucial. And

16:58

the, in response to

17:01

questions from the board in the Brazilian

17:03

decision, meta said that

17:05

it doesn't have any particular metrics for

17:07

measuring the success of its election integrity

17:10

efforts, but it pledged

17:12

to provide

17:15

the company with relevant data and to come

17:17

up with policies

17:19

for determining actions

17:21

and elections moving forward. What are the

17:24

kind of policies that the

17:26

board thinks that the company should adopt?

17:28

And Kenji, what kind of issues moving

17:31

forward in elections does the Brazilian decision

17:33

flag? Yeah, you

17:35

know, I think part of it is, you

17:37

know, holistic. And you can get the same

17:39

thing, the election itself is the kind

17:42

of aftermath of the election. And

17:46

making sure that the team

17:48

doesn't just dissolve and meta doesn't take

17:50

its eye off the ball. So making

17:53

sure that crises and

17:56

concerns are tailored to the extent

17:58

of the crisis. crisis. But

18:01

also, you know, skipping ahead a little bit,

18:03

you know, one of the conversations that we

18:05

had more recently was with regard to manipulated

18:07

media, and the Biden case

18:10

where Biden was portrayed

18:12

through manipulated media as inappropriately

18:14

touching, you know, his adult

18:17

granddaughter's class. And

18:19

so this was, you know, making

18:22

a round and was challenged under

18:24

meta-symmetal media policy. And interestingly, the

18:26

board, you know, kept it up,

18:29

because we said, as we apply

18:31

the manipulated media standard, it doesn't

18:33

actually cover the instance. So the

18:36

manipulated media standards said, you know,

18:38

you cannot make people say things

18:40

that they didn't actually say.

18:43

So, you know, Nancy Pelosi, that famous

18:45

video of her slurring her words, for

18:47

example, might have been covered

18:49

under that. But somebody doing something that they

18:52

didn't do here, you know, touching your granddaughters

18:55

breast inappropriately, was not

18:57

covered under the strict kind of four corners of

18:59

the policy. Similarly, the manipulated

19:02

media policy as written at the time

19:04

said, you know, look, we're not going

19:06

to actually deal with manipulated

19:09

media acceptance of far as it concerns

19:11

AI. And this did not, it was

19:13

kind of just a very pedestrian usage

19:16

and manipulation of media

19:18

and did not involve AI.

19:20

So it just fell outside

19:22

of the policy. So this is one of the

19:24

instances where we say you actually reached the right

19:26

decision. But the fact that this is the right

19:29

decision means that the policy itself is wrong.

19:31

So the community standard was something that

19:33

we urged, you know, meta to

19:35

review. So that was actually the

19:38

more dramatic, you know, instance where

19:40

we said, meta, you know, please

19:42

clean this up and please clean

19:44

this up before the, you know,

19:46

upcoming elections in the

19:48

span or election year. And the

19:50

response on meta's part was

19:52

very swift. Michael,

19:55

tell us about what

19:57

that response was and what the

19:59

Biden video case tells

20:02

you about whether the meta-standards themselves

20:05

track the First Amendment closely

20:07

enough or not. The distinction

20:09

between saying things you didn't

20:12

say and not doing things you didn't do isn't

20:14

intuitive. How did

20:16

the board, how

20:18

did meta-respond and what

20:20

do you think meta-should

20:22

adopt as a policy for

20:24

manipulated media? As

20:27

Kenji said, the manipulated media standard which

20:29

was adopted I believe in 20 or

20:31

20 or several

20:34

years ago, which is eons

20:37

ago in terms of the

20:39

actual development of the technology,

20:42

had to do with what they

20:44

were seeing at the time. It

20:48

was easier at the time to

20:51

fake voices and

20:53

statements and that was basically

20:55

the problem but things change

20:57

and meta I

21:00

think to their credit immediately adopted,

21:03

we identified three different distinctions within

21:05

the policy that we thought no

21:07

longer, if they ever made sense,

21:10

they don't make any sense anymore

21:12

and they immediately moved toward

21:17

correcting those and

21:19

in addition to that, there was one

21:22

other point which is that under

21:24

the prior policy meta either if

21:28

it came within the policy they would take

21:30

the post down or if it

21:33

didn't they would leave it up

21:35

so it was a binary leave up take down

21:38

sort of thing and it is our

21:40

judgment that particularly in the

21:42

area of manipulated media that

21:45

the real problem has

21:47

to do with the fact that readers,

21:50

users, viewers are frequently

21:53

unable to tell whether something

21:55

has been manipulated or not

21:57

and that's the deception. It

22:00

isn't so much the content. I

22:02

mean, in politics, people say falsehoods

22:04

all the time, right? And

22:07

especially about candidates. And

22:09

our United States Supreme Court has been

22:11

very strong in protecting

22:14

even falsehoods in the connection

22:16

with political speech. But

22:21

what we concluded was that it was

22:24

most important for meta

22:26

to inform readers,

22:29

users, viewers when

22:32

the manipulation has taken place

22:34

so that we would

22:36

not be deceived into thinking this

22:38

is real. Because some of, especially

22:41

as AI gets better and better,

22:43

the deceptions are harder and harder

22:45

for us to detect. This

22:49

is so interesting

22:51

and nuanced,

22:53

the content policy that you

22:55

recommended in the Biden

22:58

decision, as you just

23:00

said, Michael, includes three recommendations to

23:02

address the harms posed by manipulative

23:04

media. Meta should

23:07

reconsider the scope of the policy

23:09

to cover first audio and audio-visual

23:11

content, two content showing people doing

23:14

things they didn't do as well as saying things they

23:16

didn't say, and three content regardless of the method. You

23:19

also ensured, asked for a

23:22

clearly defining in a single unified

23:24

policy the harms that meta sought

23:26

to prevent beyond preventing users from

23:28

being misled. And third, you

23:30

asked for proportionality and said meta should stop

23:32

removing manipulated media when no other policy

23:35

violation is present. It took me a while

23:37

to read all those recommendations. But Kenji, reflect

23:39

on the dynamic of

23:41

the board, really in a

23:44

thoughtful way, weighing First

23:46

Amendment values, recommending that the

23:49

company incorporate them in policy.

23:53

Do you feel that free

23:56

speech values are being well respected, and

23:58

how is this dynamic? working? Yeah,

24:02

well, I mean, it's not really

24:04

free to each values alone, right?

24:06

Because, as meta

24:09

says, and as we deeply

24:12

believe, voice is not

24:14

the only concern that

24:17

we need to protect. So there are

24:19

really five values that meta articulates as being

24:22

salient to its platforms. And from time to

24:24

time, it's just freedom of expression is paramount,

24:26

right? So we might leave with voice as

24:28

you have. But there are

24:30

four other voices too. There's values

24:32

as well. There's privacy, there's authenticity,

24:35

there's dignity, and there's safety. So

24:37

what I find to be incredibly

24:39

fascinating about my time

24:41

on the board is that the

24:44

balances among these different values can

24:46

be extremely tricky. And that in

24:48

many ways, we are kind of

24:50

not as tilted over one wing, vis-a-vis

24:53

speech. I'm curious as

24:55

to whether Michael agrees or not,

24:57

as US First Amendment jurisprudence would

25:00

indicate. So I've written a

25:02

little piece on this, and it's pushed me

25:04

back into the history of the creation of

25:07

the oversight board. And in

25:09

fact, early on, with regard to content

25:11

moderation, long before the oversight board even

25:13

existed, content moderation was done

25:15

by First Amendment standards

25:17

that the people who were in charge

25:19

of these issues were trained

25:22

in the US First Amendment tradition

25:24

of yes, the Nazis get to march in Skokie,

25:26

yes, you get to burn across on your

25:29

yard because the very both of our

25:31

First Amendment jurisprudence, as the Supreme Court

25:33

recently reminded us in the Mattel vs.

25:35

Tom case, is that we protect even

25:37

the speech that we abhor. And

25:40

so the speech we hate doctrine

25:42

is part of a very expansionist

25:44

and rigid vision that is intensely

25:47

speech protected. And the

25:49

story goes with meta is that as meta

25:51

became more global, realize what an outlier the

25:53

United States was and could not simply default

25:55

back to US First Amendment

25:58

jurisprudence because, you know, If

26:00

you go to

26:02

Europe and you look at sort of

26:04

hate speech laws there, the cast is

26:06

very, very different and much more tilted

26:09

towards equality or dignity than it might

26:11

be towards speech. And so

26:13

as a global platform that insisted on

26:15

having community standards that were not geofence

26:18

that applied across the globe, the

26:20

balance, I think, has to be struck

26:22

differently. And that's why our baseline here

26:24

is not the U.S. Constitution and free

26:27

speech, but rather international human rights norms.

26:30

So one of the things that I

26:32

greatly appreciated on being on the board

26:34

is that because we're not state actors

26:36

and because we are global, we

26:38

can strike these balances differently from

26:41

what I was taught and what I teach

26:43

are the First Amendment norms with regard to

26:45

free speech. And sometimes we acknowledge that in

26:48

our decisions. So I'm adventuring a little bit

26:50

further afield, but I still hope this is

26:52

useful. We issued a

26:54

decision on blackface where we said, if

26:56

you actually portray yourself in blackface, even

26:59

if it comes from an

27:01

innocent tradition like the Dutch

27:03

Black Pete tradition, we are

27:05

going to prohibit it. And

27:07

we know that if we were a state actor, if

27:10

not a were a state actor, we would not be

27:12

able to do that. Right. International

27:14

human rights law would probably not allow that,

27:16

you know, absent some exception like violence or truth

27:18

threats or something like that. But

27:21

given that we are not regulating

27:23

a state actor, given this is a private,

27:25

you know, kind of super compliant to the

27:27

situation, we can strike that balance differently and

27:29

on balance, you know, equality norms, or dignity

27:32

norms, you know, speech norms in this particular

27:34

hate speech context. So that's a hate speech

27:36

case, but all of these issues go to

27:38

what the board is trying to do in

27:41

elections as well, which is

27:43

to derive and balance out these different

27:45

values. And what's been really striking to

27:47

me is that, you know, if the

27:49

baseline is international human rights norms, oftentimes

27:51

that calculus comes out differently than it

27:53

would if the baseline were US First

27:55

Amendment norms. Fascinating. Michael,

27:57

you are a... a

28:00

prominent defender of US First

28:03

Amendment norms, do you feel

28:05

that the board is respecting

28:08

them sufficiently in using international human

28:10

rights as a baseline or

28:12

would you strike the balance in a different place? So

28:16

I've come to a

28:18

similar place to Kenji,

28:20

but I think

28:22

the important difference is not so

28:24

much the difference between

28:26

international and especially

28:29

European norms versus

28:31

US norms. In fact, if I

28:33

thought the project of the board

28:35

were to bring European understandings of

28:40

the balance between expression and other

28:42

things to the US, I

28:44

would be an unhappy person. I

28:48

think the more important point is the one

28:50

that Kenji was making toward the end, which

28:52

is the difference between a private

28:54

company and a government.

28:57

So even within the United States, private

29:00

companies are free to

29:02

not convey speech

29:09

that they disagree with over their platforms.

29:11

Newspapers can do this and there

29:13

are actually, I think three

29:16

cases before the Supreme Court this

29:18

year talking about

29:20

that balance. And I think

29:23

when you reflect upon this, that

29:26

that's what almost anyone would recognize.

29:29

So take for example, nudity

29:31

and pornography. Our

29:35

constitutional law does not allow

29:37

the state to criminalize

29:40

a lot of erotica,

29:45

falling short of obscenity,

29:47

which is fairly strictly defined.

29:51

Meta made a choice very early

29:53

on, maybe at the very beginning

29:55

of the platform, that it

29:57

was going to be a family friendly.

30:00

platform and was not going to allow

30:02

it to be taken over as some

30:04

platforms have been by

30:07

a whole bunch of sexual

30:09

garbage. Now the state

30:12

couldn't do that, but META is

30:14

free to do that. And

30:18

spam is another thing. The government

30:20

couldn't make spam illegal, but certainly

30:23

the platforms have a right

30:25

to keep spam off. META

30:28

also made a decision in

30:30

favor of authenticity, that is people

30:32

speak in their own voice. Well,

30:35

out in the constitutional

30:37

world of the First Amendment, people

30:39

very frequently have a right to

30:42

speak anonymously or

30:44

pseudonymously. After all, the

30:46

Federalist Papers were published

30:48

under the name of Publius,

30:51

not under the actual author's

30:54

name. So all of these are

30:56

actual differences between private companies and

30:58

the government that we

31:02

constantly are taking into consideration

31:04

and frankly having to figure

31:06

out because although we

31:08

refer to international human rights decisions,

31:10

most of those are also about

31:13

governments. And it's our

31:15

task to try to figure

31:17

out to what extent that properly applies

31:20

here. Now we started out talking

31:22

about the manipulated

31:25

media standard. And here I think

31:27

we have landed much

31:30

more closely where the US Supreme Court

31:32

has on freedom of speech. That is

31:34

to say falsehoods can

31:38

only be banned as such when

31:41

there's an identifiable harm from them.

31:44

That's the reason why we

31:46

say don't take it down just

31:48

because it's manipulated media. Give

31:51

the viewers the

31:54

information they need to be able

31:56

to tell whether they're

31:58

being manipulated or not. but don't

32:00

take it down unless it violates some

32:02

other independent standard,

32:04

which is to say it produces

32:06

some other harm. And that

32:09

is, frankly, just exactly the holding

32:12

of the United States in a case called

32:14

the United States versus Alvarez. Wow.

32:18

Kenji, tell us

32:20

more about whether you think the board is striking

32:22

this balance between speech and other norms like dignity

32:24

and privacy in a thoughtful

32:26

way among the cases that the board

32:28

relied on in the manipulated media case

32:31

was one involving sexual

32:33

harassment in India where the board have held

32:35

Metta's decision to restore a post

32:37

to Instagram containing a video of a woman

32:39

being sexually assaulted by a group of men.

32:42

The board found that Metta's newsworthiness

32:45

and our allowance is inadequate in resolving

32:47

cases like this, and the company should

32:50

introduce an exception to the adult sexual

32:52

exploitation policy. It's just another example

32:54

of how, in what

32:56

a nuanced way, that the board is

32:59

striking these balances as you step back

33:01

and look at the board, how

33:04

do you think it's doing? Yeah,

33:06

I am a true believer here.

33:08

So I think the board is doing an excellent

33:11

job. And if anything,

33:13

I think that this model could be

33:15

exported to other contexts. So just

33:18

to clarify where I was

33:20

going with the European

33:23

standard, that's really a story,

33:26

as I understand it, of how

33:28

the standard evolved within Metta itself. So

33:30

it's the origin story rather than a

33:33

normative story about what

33:35

Metta's strongest defense is with regard to

33:37

how it would balance norms differently

33:40

than first amendment jurisprudence would. I

33:42

think the normative story is very much the

33:45

private public distinction that Michael

33:47

has also fastened on. But if you think about

33:49

that for a minute, if it can pan out,

33:52

that applies to other contexts as well. So

33:55

I'd like to look back to the kind of

33:57

ill-fated congressional

33:59

hearings on air. anti-Semitism on the

34:01

Hill. The University

34:03

of Pennsylvania president kept saying, I

34:05

adhere to the Constitution, we adhere

34:07

to the Constitution's First Amendment kind

34:11

of protections in response to whether

34:13

calls for genocide would incur some

34:15

punishment on U Penn's campuses. And

34:18

she got accused of being overly

34:20

legalistic in her answers and being

34:23

kind of bloodless or overly analytical

34:25

in the way that she responded.

34:28

I actually had a different concern. I was

34:30

worried that she was not being legalistic enough

34:32

because the Constitution

34:35

does not bind the University of Pennsylvania. So

34:37

to act as if U Penn was

34:40

bound by the Constitution anyway or free speech

34:42

norms in any way, I think was wrong,

34:45

right? Or it just is wrong, right? And

34:47

she was, she's very, very smart. She's a

34:49

constitutional scholar herself. Like she herself was very

34:51

clearly aware that she wasn't bound, but

34:54

she was kind of wrapping herself in the Constitution

34:56

in a way that I thought was ill-advised

34:59

because she could

35:01

make the choice and Penn could make the choice

35:03

to be bound by the Constitution's norm. But

35:06

to ally the fact that it is a choice, I think

35:08

is a mistake. Because I think

35:10

increasingly private actors, whether they

35:12

be universities, whether they be companies, are

35:14

all gonna be called upon to make

35:17

these very thorny decisions about speech.

35:20

And I don't think that they're gonna simply

35:22

be able to bleed out the First Amendment

35:24

free speech and expect the United States Supreme

35:27

Court to do that. So it raises the

35:29

question of why so many individuals do that.

35:31

And part of it may be that people are

35:33

reading free speech broadly as a norm rather than

35:35

narrowly as a rule. But I think the main

35:37

reason people do that is that these speech issues

35:41

are so difficult and so painful that it's

35:43

much easier to just outsource them to the

35:45

United States Supreme Court rather than to deal

35:47

with them on their own. So love it

35:49

or hate it. One thing that I really

35:52

do admire about meta and the OSB is

35:54

that they at least have the courage to

35:56

say, we don't think that the United States

35:58

Supreme Court can be a guide, we

36:00

don't think that we ourselves can do it

36:02

on our own, and so therefore we want

36:04

a separate independent body to do this. So

36:07

when I think about this in the university

36:09

context, whether that's NYU or Stanford or UPenn,

36:11

I do think ultimately they are going to

36:13

have to create some version of the OSB.

36:16

And what are the components of that? I

36:18

think it's going to be a credible body

36:20

of individuals who are independent, a credible body

36:23

of law, so whether that's your own university

36:25

standards. Ideally, it wouldn't be something that the

36:27

university could just change at will. So ideally

36:29

something more like international human

36:31

rights law that stands independent. It could

36:34

be AUP principles or what

36:36

have you, but so long as the university

36:38

can't unilaterally change those principles. And

36:40

then finally, public reasoned opinions, because

36:43

one of the things that Feldman

36:45

is often credited with driving the

36:47

creation of the OSB,

36:49

at least in part, said was people

36:51

are going to disagree in good faith

36:53

about correct or incorrect decisions. So the

36:55

greatest kind of honor that

36:57

you can do to the fact that

36:59

they're dissenting views on this subject is

37:01

to give public reasoned explanations for what

37:04

you're doing. The universities in general that

37:06

I've looked at do not have those

37:08

open transparent procedures of the credible body,

37:10

interpreting a credible body of law with public

37:12

reasoned opinions. I think that those three elements are

37:14

the core of what the OSB

37:16

does, and I think what makes the OSB

37:19

a noble experiment. Reasonable

37:22

body of law, international human rights

37:24

norms, and public reasoned opinions. Michael,

37:26

do you agree that

37:28

the oversight board might

37:31

be a model for universities? And tell us

37:33

how the oversight board has dealt with

37:36

some of the issues that are now

37:38

convulsing universities, including its

37:40

decision to take expedited cases

37:42

on speech and videos about

37:44

Gaza. Yes,

37:48

I mean obviously the

37:51

board is about social

37:54

media, not about universities,

37:56

but there was at Stanford there

37:58

was a a group

38:00

of law school

38:03

people, including students who, you

38:05

know, five or six years ago, you

38:07

know, before the universities

38:09

were convulsed with the, with

38:11

the really awful things

38:14

that are happening right now, recommended

38:17

that the university have

38:19

a free speech ombudsman. This

38:22

is, this would be not quite a

38:24

board, but it would be someone who

38:26

is, stands outside

38:28

of the needs,

38:30

ordinary day-to-day needs of the

38:33

university and specializes in, like,

38:36

would be chosen specifically for the purpose

38:38

of thinking more

38:40

in a more balanced way about

38:44

freedom of speech in the university context.

38:46

So that would be somewhat

38:48

similar to the board. Also

38:50

our board just, we are only

38:53

speaking directly to

38:56

Meta, but there are other social

38:58

media companies out there as well. And

39:02

now we think, and I think we have sort of

39:05

reason to believe through, you

39:07

know, rumor channels and that

39:09

sort of thing, that some

39:11

of the other companies actually read our opinions and

39:14

that we may be more influential than, than it

39:17

appears on the surface. Because of

39:19

course, you know, other companies are

39:22

not going to say that Meta's oversight

39:24

board is influencing

39:26

their policies, but to

39:28

the extent that we

39:30

are offering, I think

39:32

thoughtful answers to some of these questions,

39:36

I think it could spread. By the

39:38

way, you've got Kenji and me on here,

39:41

you know, to talk about what we

39:43

think, you know, maybe you should have

39:45

another program in which you invite critics of

39:47

the board because they exist too. And,

39:50

you know, they have something to say.

39:52

I'm not going to tell you that

39:54

every single decision of the board has

39:56

been right. I've disagreed

39:58

with some of them. I

40:00

do think that the general direction has

40:02

been excellent. And

40:05

I think any powerful institution

40:08

benefits by having independent people

40:10

looking over their shoulder. It's

40:13

not just what the people looking over

40:16

their shoulder actually say or do. It's

40:18

the fact that they are looking over

40:20

their shoulder just brings more caution

40:24

and thoughtfulness to the process. Well,

40:27

we would love to have more programs on

40:30

this incredibly illuminating

40:32

topic. Kenji, you've talked about the utility of

40:35

the boards combining left and right cases on

40:37

the same issues, like the board's decision to

40:39

bundle abortion speech cases on the left and

40:41

the right. And you wish that

40:43

the Supreme Court could do the same thing instead

40:45

of hearing single cases. Tell

40:49

us about the expedited speech cases

40:51

involving videos about Gaza.

40:55

What issues do they raise and how does

40:57

the board deal with them? Yeah,

41:00

so sometimes just by

41:02

way of quick background, we do

41:05

formally bundle cases that are on

41:07

a similar topic. And as you

41:09

noted, with abortion speech, both

41:12

pro-life and pro-choice, we bundled

41:15

instances of that speech and then protected

41:17

it across the board. So this is

41:20

actually a very useful way, I think,

41:22

to make sure that we

41:24

keep the stakes of the cases suddenly

41:26

visible without it being about whose ox

41:28

is getting gored or whose party is

41:30

being favored or disfavored. So I just

41:33

want to harken back to an important

41:35

thing Michael said earlier, which is that

41:37

the Trump case wasn't just about Trump.

41:39

In fact, it's a fact to extend

41:42

far beyond Trump. And it's really important

41:44

to keep that as legible as possible. And

41:46

one of the ways in which we do that

41:48

on the board is to take speech

41:51

from opposite sides of the spectrum to link

41:53

them together and then let the ideologies cancel

41:55

each other out while reaching

41:57

because we reached the same decision.

42:00

on both in a speech-protective vein. So

42:03

the expedited cases, which we just started taking

42:05

this past year, well,

42:09

in 2023, had to do with the conflict in

42:14

the Middle East. And what we

42:16

did was not to formally bundle,

42:18

but to link two cases. And

42:21

the two cases both concerned quite

42:23

harrowing, viscerally kind of moving

42:25

videos from both sides. So there was one

42:28

from Gaza City, a hospital

42:31

there, that had

42:33

just been destroyed. And

42:35

then there was some very poignant,

42:38

powerful scenes of

42:40

the suffering in that hospital.

42:43

And then there was another video that came

42:45

from, that depicted

42:47

a scene from Israel where a woman

42:49

was pleading with her

42:52

kidnappers not to be kidnapped by them.

42:55

And so the notion was, yeah, we

42:57

understand why the impulse might be, no

43:00

one should have to see this. It's

43:02

an incredibly volatile period, incredibly sensitive time.

43:04

But we felt that whether

43:06

we protected or didn't protect the

43:09

videos that are served, if

43:11

they were linked to each other, because what we

43:13

ended up doing was to say both videos should

43:15

be left up. People in the

43:18

hospitals should have the capacity to know that

43:21

the relatives can find them. People who are

43:23

family and friends of the individual being kidnapped

43:25

should be able to mount a search for

43:28

that individual. There are

43:30

really strong humanitarian reasons on both

43:32

sides that without saying that this

43:34

is not incredibly difficult, so

43:39

seemingly intractable conflict in the Middle

43:41

East, this was kind of

43:43

our piece of it to say, we're gonna

43:46

protect each on both sides and we're gonna

43:48

link these two cases so that people can

43:50

understand why we did that without letting ideology

43:53

get too much in the way. Fascinating.

43:56

Michael, how do you think the board

43:58

is dealing with Gaza? like issues

44:01

could one post videos

44:04

from the river to the sea the

44:06

kind that became of controversy

44:08

during the hearing with

44:10

the college presidents and does the

44:12

boards approach to the Gaza

44:14

conflict Teach anything to universities

44:17

who are dealing with the conflict today So

44:20

we these first two cases were expedited.

44:22

They were decided and I think it

44:24

was about a two-week timeframe

44:28

which They were the first

44:30

ones we dealt with under

44:32

these procedures I mean if I think the

44:35

biggest legitimate complaint about the

44:37

board is just the ordinarily our

44:39

work is too slow because of

44:41

the way we

44:43

go about things but We

44:45

did this I think on a rather

44:48

rapid timeframe, but since then there have

44:50

been other related matters That

44:52

come up, you know specific cases and

44:56

also We

44:59

we reviewed the Very

45:02

touchy question. I don't think people in the

45:04

United States are even aware of which

45:07

is the word Shaheed in

45:12

Arabic Arabic means It

45:15

has a range of meanings. It refers

45:17

to people who've you know recently been

45:19

or have been killed it

45:22

sometimes means martyr and it is

45:24

often used in connection with Terrorists

45:28

who are who die in the course of

45:30

trying to do that and that's that's it

45:32

touches on that issue we were talking about

45:35

before a praise for tougher

45:37

terrorists or other dangerous organizations, but it

45:39

has a much broader use than that

45:41

it can be if a Teenager

45:45

dies in an automobile accident They

45:48

might well be referred to as

45:51

Shaheed by their family and so

45:53

forth and Shaheed was the single

45:55

word That on

45:57

most often led to take

46:00

downs of messages. And

46:02

it had a really

46:06

weird and sometimes

46:08

counterproductive effect as

46:11

applied in the

46:13

Middle East. Again, in the United States,

46:15

we were kind of oblivious to that.

46:18

But we issued a policy

46:22

recommendation that

46:25

drew much closer lines to try to

46:27

make sure that the word Shaheed is

46:30

being disfavored only

46:32

in context where it

46:35

should be and not so

46:37

broadly. Now, Jeff, you ask about from the

46:40

river to the sea. That

46:42

happens to be, there's been a public announcement

46:44

that happens to be a specific case

46:47

that is being taken up by the board. And

46:49

I'm not going to predict how it's going to

46:52

come out. But I

46:57

may be open. I can't remember

46:59

exactly the timeline, but your listeners

47:01

may be interested to know that

47:03

we announce cases when they're being

47:05

taken up and we welcome

47:08

public comment. A lot of

47:10

those comments are from civil society

47:12

groups of various sorts and academics who

47:14

study the matter. But anyone is

47:18

free to comment. And

47:20

we take those all

47:23

into account and very seriously. So,

47:26

cases like this, some

47:29

of the listeners on this program may want to chime

47:32

in. That

47:35

would be very constructive. And we will

47:37

post the links to allow people to

47:39

do that. Back

47:41

to election integrity, the board has

47:43

decided important cases recently involving

47:46

a Cambodian prime minister. It overturned

47:48

Metta's decision to lead up a

47:50

video on Facebook in which the

47:52

Cambodian prime minister, who Sen threatened

47:54

his political opponents with

47:58

violence. And

48:01

there was also a case

48:03

involving reporting on the

48:06

Pakistani parliament, which involved the Oversight Board

48:08

upheld now, met its decision to lead

48:10

up a post shared by a news

48:12

outlet in Pakistan that includes a video

48:14

of a politician giving a speech to

48:16

the country's parliament. The board

48:18

considers that safeguarding such figurative speech in

48:20

a run-up to elections is fundamental. Kenji,

48:22

tell us about these cases. Yeah,

48:25

they're actually really good cases to

48:27

consider together. So the

48:29

Hun Sen case led to actually the

48:31

entire Oversight Board being banned from Kabul

48:34

to Cambodia because we were deemed kind of

48:36

persona non grata. So I suppose we can

48:38

wear that as, you know, a

48:40

badge of pride for having some effects there.

48:42

But in that case, the

48:46

Cambodian Prime Minister incited violence against his

48:48

political opponents. This is now the father

48:50

of the current Prime Minister because there's

48:52

been a regime change. But

48:54

the Cambodian Prime Minister said we will go

48:56

after our opponents with the bat, right,

48:59

and incited for the violence. And that

49:01

way, we deem that to be not

49:05

permissible. We told

49:07

men about any who's getting it now.

49:10

And then we also urged this

49:12

extension of the Prime Minister's

49:14

account. You know, Pakistani

49:16

case was, you know, on the surface,

49:18

perhaps similar, right, it was an individual

49:21

in Pakistani, you know, parliament, an elected

49:23

official making a speech saying

49:25

that, you know, individuals should be hanged, right. And

49:27

so that might seem to be even more

49:30

serious. But when we looked at the

49:32

context, you know, it was much

49:34

more figurative, it was similar to our,

49:37

you know, Dasta Khomeini case,

49:39

which was, we saw as kind of down

49:41

with Khomeini. But one

49:43

of the salient distinctions among those cases

49:46

is that the media environment

49:48

was very, very different in Cambodia, where

49:51

Hun Sen, you know, kept the Prime

49:53

Minister at issue in that in that

49:55

case, had a very,

49:57

very long history of suppressing a tradition.

50:00

sources of media. And so in that

50:02

media environment, his defense was, you know,

50:04

this should be, and

50:07

men's defense was, this should be politicized because

50:09

it's newsworthy, right? And that just seemed a

50:11

little bit backwards to us. It seemed like

50:13

somebody was taking advantage of first shutting down

50:15

any other source of media other than themselves

50:17

and then saying that their Facebook feed should

50:19

be kept out because they

50:21

were newsworthy. So this to me struck,

50:24

you know, that just had the kind

50:26

of tincture of that old law school

50:28

saw that you can't sort of kill

50:30

your parents and then throw yourself on, you know,

50:32

after coming from the court on the ground that

50:34

you're an orphan, right? I mean, you cause the

50:37

situation that you're now trying to, you know, take

50:39

advantage of. And so, you know, one

50:41

of the things that strikes me in

50:43

these, in these cases that I frankly

50:45

wasn't anticipating as much with how profoundly

50:47

context dependent each one of them, you

50:49

know, would be, right? So those,

50:52

you know, cases sort of rabbited off

50:54

in different directions, even though they seem

50:56

like they are concerned, you know, relatively

50:58

similar forms of speech because of distinctions

51:01

pertaining to media, but also, you know,

51:03

others that, that made

51:06

the difference between them. Can

51:08

I just add one more detail about the

51:11

Cambodian situation? I think this is an

51:14

example of where having

51:16

an independent board like the oversight

51:18

board really helps

51:20

meta because they, because

51:23

Hunson's immediate reaction was

51:26

to shut

51:28

down meta, shut down Facebook

51:31

as the important, more important

51:33

platform there, shut Facebook down

51:35

entirely. And of course

51:37

they don't like that for commercial reasons, but

51:39

they also don't like that because Facebook is

51:42

such an important medium of

51:45

communication among Cambodian people, not

51:47

just for political speech, but

51:49

even for, you know, very

51:51

ordinary communication. And

51:53

to have the whole country

51:55

deprived of the platform was,

51:59

it was a very, heavy weapon

52:02

on Hundson's part. But

52:05

because the, you know,

52:08

Matt can say, oh, but we

52:10

are committed to following the oversight

52:12

board's decisions, they can sort of

52:14

lay, they can outsource

52:16

the blame for this. And

52:18

so instead he banned the 22

52:21

members of the board from

52:23

being able to come to Cambodia rather

52:25

than shutting down Facebook

52:27

altogether. And I think that

52:30

there would not have been that kind of good

52:33

cop, bad cop benefit in

52:36

the absence of something like

52:38

the board. Now,

52:41

I should point out this, I

52:43

think is a little disappointing that

52:45

this is the company

52:48

did not fully follow the board's

52:52

recommendations here. I

52:54

think it's probably the only really

52:56

important recommendation in the history of

52:58

the board where they flatly said

53:00

we cannot do part of it.

53:03

And that had to do with the

53:05

future treatment of Hundson's accounts

53:09

and over the platform. And

53:12

can I slightly pull against that, which is to

53:14

say, you know, good cop, bad cop

53:16

to me, Michael suggests that like it's the same

53:18

kind of entity, right? Deploying

53:20

two different tactics to achieve the same end.

53:24

I think that kind of scans the fact

53:26

that the reason that there's conflict as evidenced

53:28

by the fact that they didn't sort of

53:30

go meta did not adhere to the suspension,

53:32

that we are really independent, that we come

53:34

to different assessments, that they're looking at the

53:37

world through in their tooling titles to do

53:39

this from a profit maximization viewpoint, we're looking

53:41

at it from a human rights viewpoint. So

53:43

perforce, we're going to come to different answers.

53:46

And I think that's a really important question. Just so there's no

53:48

confusion among the viewers and listeners today

53:51

about what's mandatory and what's not. Whenever we

53:53

have a takedown leave up decision, meta has

53:55

to adhere by that that is mandatory and

53:57

they have never flouted that. recommendations,

54:00

they have to respond, but they do not

54:02

have to take the recommendation.

54:04

In this particular case, as Michael said,

54:07

you know they did not. But you know this

54:09

right of response that we have I think

54:11

is critically important because it's what I'm beginning

54:13

to understand now having been you know on

54:15

the board, I'm the newest member of the

54:17

board, so I've been on for a year

54:19

and change now. It's how iterative you know

54:21

the conversation is. Inevitably there'll be another case

54:23

that raises the same issues and then we'll

54:25

be right back into the conversation again.

54:28

So the fact that we give recommendations and then

54:30

they have to speak back to us and then

54:32

we get to speak back to them again next

54:34

time similar issues arise means that we're

54:36

in this you know conversation together. But

54:38

it's conversations that are being approached you

54:40

know through positions of complete independence from

54:43

each other rather than you know

54:45

trying to voice you know the same strategy in

54:48

two different ways. As

54:50

a final thoughts

54:52

on the Cambodian case and then we'll

54:54

have closing thoughts. I know you wanted to flag how

54:56

people often use the pretense of misinformation

54:58

to take stuff down and how quick governments are

55:00

to latch on to this claim of misinformation

55:03

when they're trying to avoid bad publicity. Is that

55:05

one of the dynamics here? I

55:08

think it is. I mean in the Hun Sen

55:10

case it was a threat of violence rather than

55:12

an attempt to get the company

55:16

to take down alleged

55:18

misinformation. But there are

55:20

governments, China is one of them

55:22

that tremendous effort

55:24

into getting criticism

55:26

of the government eliminated

55:29

from a world discussion

55:31

and they bring pressure in various

55:33

ways and we try to be

55:36

alert to that

55:38

and and be a

55:40

guard against it. And one thing that we've

55:42

done and this I think is a really

55:45

interesting area is it's come up in the

55:47

US Supreme Court again in a

55:49

couple of cases this year and we don't know

55:51

and that's you know what is the line

55:54

between a government for giving

55:57

sharing legitimate information with the

55:59

social social media platform

56:01

that's relevant to their taking

56:03

stuff down versus an illegitimate

56:05

use of

56:08

the threat of regulatory

56:10

retaliation to get, to

56:14

induce the company to

56:16

take down material unfavorable

56:18

to the government.

56:21

It's a line, somewhat easy

56:24

to state in theory, but

56:26

extremely difficult to police and

56:28

practice. And we have come

56:31

up with a solution to this that I

56:33

really like and hope that

56:35

it spreads. And we have

56:37

recommended that whenever a government

56:40

or government officials request

56:43

takedowns of posts

56:45

on the platform for reasons other

56:48

than legality, that is there when

56:51

the government is saying, well, this

56:53

is hate speech, or this is,

56:55

which is not illegal, or this

56:57

is misleading, information,

57:00

or that sort of

57:02

thing, that the company

57:05

simply make a public of announcement,

57:08

a list of the contacts from

57:10

the government urging that. Much of

57:13

what the government urges is entirely

57:15

legitimate, and making it public is

57:17

not going to hurt anything. But

57:19

there are times when governments, and

57:21

remember, we're talking about governments all over

57:24

the world, but the United

57:27

States government is not immune either. But

57:30

there are times when they

57:32

are misusing their

57:37

power over the companies

57:39

to avert criticism. And

57:42

so this is an area where a little

57:44

sunshine would help

57:48

the problem. The United States Supreme Court

57:50

doesn't have that option. It has to

57:53

decide the line about when the government

57:55

goes too far and when it doesn't.

57:57

But we and META

57:59

have a. The chance of an

58:01

era of of the solution there

58:03

is a less binary than that.

58:07

Was. The that so superb node

58:09

said. Emphasize! At the end

58:11

of this great discussion I have

58:13

to say how. Scott.

58:16

Our house has struck. I am by that

58:18

but a nuance and thoughtfulness of these discussions

58:20

and also by the. Fact

58:22

that both of you coming from different

58:24

perspectives really do find yourself substantively a

58:26

green. Thumb. On these

58:29

complicated issues are more often

58:31

than not. are you both

58:33

like the importance of transparency?

58:36

And accountability as one of the factors that's

58:38

made that possible. The kenji as you reflect

58:40

on. Why? The board is

58:42

able to have these civil discussions, insults, and

58:45

to tend to agree more often than not

58:47

of what are some of the factors that

58:49

point you. Just.

58:51

Add to come to mind so one

58:54

does us last credit and the other

58:56

does more as far as as a

58:58

second one more or the first ones

59:00

I think free speech made strange bedfellows

59:02

of on politics. I saw that as

59:04

people who are on from different sides

59:06

of the ideological spectrum are. Not

59:08

micron. I. I like to think that I'm a

59:10

reasonable forgot the other Michael. The region will come

59:12

from it over and I'll be are I agree

59:14

Com at the fun with the for an ideological

59:17

pirate them find one of them to. This is

59:19

the free speech. And often times that

59:21

a little malibu right notes on you actually

59:23

that's a political candidate make. It

59:25

out and buy them in a utterances

59:28

not gonna take about the or political

59:30

opponent by the motherboard was. A

59:32

unanimous and and are behind that

59:34

the satellite service that lion idea

59:36

of like business. Strange Bedfellows. The.

59:39

Other one I hope that more credit to

59:41

have which is that in a we are

59:43

and that kind of noble experiment together we

59:45

are you know driving towards the same and

59:47

that the work and of our inclusion of

59:49

our long enough the in our as of

59:52

I have one of the things that he

59:54

realized can around in a otherwise quite significant

59:56

differences between individuals as a common goal right

59:58

and a total commitment to that. Go

1:00:00

above and I wanted a great on Guess

1:00:02

the habit of the public eye before my

1:00:04

goal but at the far from I target

1:00:06

on the boards been able to serve web

1:00:08

people are significantly more conservative than I am

1:00:10

be have a team on that mob what

1:00:12

if issues. I've learned an incredible amount from

1:00:14

them and I know that near and dear

1:00:16

to your heart at the National Constitution Centre

1:00:18

in I just I hope that the says

1:00:20

in an experiment that you can rejoice in

1:00:22

as well. I I

1:00:25

can absolutely rejoicing and and in

1:00:27

this very inspiring. Our discussion,

1:00:29

which is a model for the kind of civil

1:00:31

dialogue the we're hoping to. Encourage Michael

1:00:33

Last words: You were a what are the

1:00:36

factors on the board? The user's account for

1:00:38

it. The Civil Dialogue and.

1:00:41

Unexpected. Agreements are that

1:00:43

is is demonstrating. Actually

1:00:45

it if if you don't mind I'd

1:00:47

just like to say something to the

1:00:50

due to the audience the my as

1:00:52

soon as is pretty diverse to the

1:00:54

or the the board will do better

1:00:56

work. If if if we get

1:00:59

more. Input from the public

1:01:01

and in several different ways. One

1:01:03

I've already mentioned much as the

1:01:05

public comments on the specific cases,

1:01:07

but remember, the case has come

1:01:10

to us from Most of our

1:01:12

case was come from user appeals

1:01:14

and I think there isn't narratives

1:01:16

out there that the board is

1:01:18

just one more sort of. In

1:01:21

a. Globalist, left leaning in

1:01:23

an organization that the doesn't care about

1:01:26

conservative speech. And I just wanna say

1:01:28

i don't think that that is at

1:01:30

all true. I don't think it's true,

1:01:32

certainly isn't true of me. I don't

1:01:35

It's not true of Kenji. If is

1:01:37

not true of Genji, it's probably not

1:01:39

to have anybody right. Best. So I

1:01:41

am an end, but it can be

1:01:44

a self fulfilling prophecy. So.

1:01:46

and i think they're been a lot

1:01:48

of complaints from the right side of

1:01:50

the spectrum the social media is silencing

1:01:52

them to which i say appeals of

1:01:54

it's cases bring them to us and

1:01:56

let us see if we can do

1:01:59

something about it Feel

1:02:01

those cases and let's see if we can do

1:02:03

something about it. Just inspiring

1:02:06

words at the end of a great

1:02:08

discussion. It was an honor to convene it

1:02:10

and it is a model for civil

1:02:12

dialogue in a polarized age. Kenji Yoshina

1:02:15

and Michael McConnell, thank you for your

1:02:17

service and thank you for educating all

1:02:19

of us at the National Constitution Center.

1:02:22

Bye. I'll convene again soon. Thank

1:02:24

you, Jeff. Thanks. This

1:02:29

program was made possible through the generous

1:02:31

support of Citizen Travelers, the nonpartisan civic

1:02:34

engagement initiative of travelers. It was streamed

1:02:36

live on April 29, 2024. Today's

1:02:40

episode was produced by Lana Ulrich, Samson

1:02:42

Mastachari and Bill Pollack, who was engineered

1:02:45

by Kevin Kilburn and Bill Pollack. Research

1:02:47

was provided by Samson Mastachari, Cooper Smith

1:02:49

and Yara Derafe. Dear

1:02:52

We The People friends, I

1:02:54

am so grateful to those of you

1:02:56

who have written asking for book plates

1:02:59

for the pursuit of happiness. How

1:03:01

classical writers on virtue inspired the lives of the founders

1:03:03

and defined America. It's been wonderful to hear from you

1:03:05

about what you think of the book and what you

1:03:08

think of the show. Please

1:03:10

email me, jrozen at

1:03:12

constitutioncenter.org, if you'd like a

1:03:14

signed book plate and to let me know what

1:03:16

you think of We The People. Please

1:03:18

recommend the show to friends, colleagues or anyone anywhere

1:03:21

who's eager for a weekly dose of

1:03:23

civil dialogue, historical

1:03:25

depth, constitutional illumination

1:03:27

and debate. And who isn't

1:03:30

eager for all of those things? Sign up

1:03:32

for the newsletter at constitutioncenter.org/connect. And always remember

1:03:34

that we're a private nonprofit. We rely on

1:03:36

the passion, the generosity, the engagement of people

1:03:39

like you who are writing in across the

1:03:41

country inspired by our nonpartisan mission. Support

1:03:44

the mission by becoming a

1:03:46

member at constitutioncenter.org/membership. Get our

1:03:48

emails, our updates and the

1:03:51

nonstop constitutional learning that my

1:03:53

colleagues and I are honored

1:03:55

to produce every week. And of course,

1:03:57

please consider giving a donation of any amount to

1:03:59

support our work. including the podcast,

1:04:02

at constitutioncenter.org/donate. On

1:04:04

behalf of the National Constitution Center, I'm

1:04:07

Jeffrey Rosen.