Good afternoon my name is François Tanguay-Renaud I'm a professor here at
Osgoode Hall Law School I'm also the co-director of the Jack and Mae
Nathanson Center on transnational human rights crime and security and it's my
pleasure this afternoon to welcome you to the 2018 'Or 'Emet lecture a few words
about the lecture the 'Or 'Emet fund was established a long time ago now in 1976
to promote the study of law in the broadest sense the fund in itself seeks
to promote through public discussion research and scholarly writing public
and professional appreciation of the significance of things like religion
ethics culture and history in the development of the legal system 'Or 'Emet
what does this mean well it means the light of truth so the lecture really
seeks to shed light on matters that perhaps deserve a bit more truth or
truth telling as it were in 2010 the Nathanson center of which
I'm co-director decided to pool its resources with the 'Or 'Emet fund to ensure
that the lecture would be delivered on an annual basis so as a result the
themes that are not explored by the by the lecture tend to relate at least
tangentially to the Mandate of the Center now this year we're delighted to
have with us Jameel Jaffer who will already be known
most likely to many of you Jameel is the inaugural director of the Knight First
Amendment Institute at Columbia University but he also previously served
as deputy legal director at the American Civil Liberties Union the ACLU where he
oversaw that organizations work on free speech privacy technology and national
security as well as international human rights
Jameel has argued civil liberties cases in multiple appeals courts in the
US as well as in the US Supreme Court and he's testified many times before the
US Congress his cases include some of the most significant post 9/11 cases
relating to national security as well civil liberties including cases
concerning the tension interrogation surveillance targeted killing and
government secrecy he actually co-led the litigation that resulted in the
publication of the Bush administration's torture memos which is a lawsuit that
the New York Times described as among the most successful in the history of
public disclosure in the United States and Jameel also led the ACLU litigation
that resulted in the publication of the Obama administration's drone memos and
most recently with the Knight's First Amendment Institute he's also litigated
groundbreaking freedom of speech cases his recent writing has appeared in
multiple publications including the New York Times the Los Angeles Times The
Guardian the nation and a Yale Law Journal forum he's an executive director
of just security which is a national security blog and he has two
best-selling books under his belt his first book administration of torture
from Washington to Abu Ghraib story and beyond
I was co-authored with Amrit Singh and was published by Columbia University
Press in 2007 and more recently the drone memos published by the new press
in the fall of 2016 Jameel is a graduate of Williams College
the University of Cambridge and Harvard Law School he has served as a law clerk
to the Honorable Amalia Pierce of the US Court of Appeals for the
Second Circuit and he's also been a law clerk to the right honourable beveler
Beverly McLaughlin Chief Justice of former Chief Justice of the Supreme
Court of Canada so I should emphasize here that despite the fact that Jameel
has built most of his career side of the border in the United States he's still
very much a Canadian and thankfully he told me yesterday that he
still identifies as such so um so and and he's actually been involved with
many organizations on this side of the border including the Canadian Civil
Liberties Association he's actually a Distinguished Fellow of the Munk school
of global affairs and public policy and many other organizations so we're really
delighted to welcome Jameel here at Osgoode Hall Law School York University
today to deliver the 2018 'Or 'Emet lecture on the topic of digital
journalism and the new public square now just a few words about format format
Jameel tells me that he's going to speak for about 30 to 40 minutes and then
we're going to open to questions from the floor for an approximate 30 to 40
minutes once again so without further ado please join me in welcoming Jameel
Jaffer thank you Thank You Professor Tanguay-Renaud
and thank you all of you for being here so I haven't lived in Toronto
for more than a quarter century which is difficult for me to say let alone
believe but I'm always delighted to be invited to speak here because I still
have many friends and family in and around Toronto it's also an honor to be
asked to deliver this particular lecture though as you can probably imagine it's
a little bit intimidating to have to deliver a lecture whose name promises
the light of truth so in the spirit of managing expectations let me just say
right up front that I can aspire to at most a faint flicker of truth through a
heavy fog that's what that's what I'll try for a few months ago The
Guardian published a remarkable story revealing that a Cambridge University
researcher had harvested as many as 50 million facebook profiles for cambridge
analytica which is a data analytics firm that was headed at the time by Steve
Bannen one of Donald Trump's key advisors and the researcher Alexander
Cogan collected the profiles with an app called this is your digital life and
through the app he paid Facebook users small amounts of money to take
personality tests and to consent to their collection to the collection of
their data for academic purposes and then Kogan turned the profiles over to
cambridge analytica which used them The Guardian said to predict and influence
American voters choices at the ballot box
The Guardian story relied in significant part on the account of Christopher Wiley
a Canadian researcher who had helped establish Cambridge analytica on and who
later worked with Kogan on the Facebook project Wiley told the Guardian quote we
exploited Facebook to harvest millions of people's profiles and we built models
to exploit what we knew about them and target their inner demons that was what
Cambridge analytica was built on he said most of you probably remember that
story you may not be familiar though with what happened the day before it was
published as the guardians editors were readying their story for print the
lawyers received a letter from Facebook and the letter threatened a lawsuit if
the Guardian went forward with the story Facebook knew that the story would
provoke disbelief and outrage and perhaps even a regulatory response by
the US Congress so it tried to quash the story with the threat of a lawsuit now
there's nothing unusual unfortunately about powerful actors threatening
litigation to preempt unflattering news coverage intimidating letters of the
kind that Facebook's lawyer sent to the Guardian are a staple of the media
sphere in the United States and in the United Kingdom and they're probably a
staple of the media sphere here in Canada too but Facebook isn't just any
powerful actor it's one of the largest corporations in the world it has more
than 2 billion users more than 200 million of them in the United States and
through its human and algorithmic decision making it has immense influence
on how its users engage with one another and with the communities around them
it's no accident that stories about political polarization and filter
bubbles an election integrity the spread of disinformation online voter
suppression in Brazil mob violence in Sri Lanka and even ethnic cleansing in
Myanmar have Facebook as their common denominator Facebook has an invisible
but powerful influence on what we would once have called the public square and
through that influence Facebook affects societies all over the world so what are
the mechanics of that influence in a new article the legal scholar Kate clonic
argues that the social media platform should be thought of as quote systems of
governance because they're now the principal regulators of speech that
takes place online through their control of the new public square the platform's
are exercising power we ordinarily associate with state actors one facet of
that power the facet that cloning explores in her article is sometimes
called content moderation by which we usually mean the determination of which
speech should be permitted on these privately controlled platforms and which
shouldn't be Facebook for example routinely removes User Content that
shows graphic violence or nudity as well as hate speech as Facebook defines it
and speech glorifying terrorism again as Facebook defines it the other major
platforms have similar policies there's an ongoing debate about the ways in
which the companies are exercising that power whether they're taking down too
much speech or too little speech or perhaps even both in an essay published
over the summer by the Hoover Institution
Daphne Keller who's a researcher at Stanford observes that people from
across the political spectrum are convinced that the platforms are
silencing speech for the wrong reasons this debate is important because
increasingly it's the platform's rather than governments that delineate the
outer limits of public discourse and also because the platform's power to
censor isn't subject to constitutional or regulatory constraint
but content moderation at least in the narrow sense of that phrase hardly
begins to describe the platform's influence the social media company's
more fundamental power over public discourse is reflected not principally
in their decisions about which speech to exclude from their platforms but in
their decisions about how to organize and structure the speech that remains
they dictate which kinds of interactions are possible and which aren't
which kinds of speech will be made more prominent and which will be suppressed
which communities will be facilitated and which will be foreclosed if the new
public square were an ocean Facebook would control not only which fish got to
swim in it but also the temperature and salinity of the water the force and
direction of the currents and the ebb and flow of the tides
but despite the rule that Facebook and other platforms now play our collective
understanding of them is very limited and we sometimes struggle even to
describe what they are when is a platform a publisher when is it a common
carrier for the past year the knight Institute that Institute that I direct
has been litigating a First Amendment challenge a free speech challenge the
president Trump's practice of blocking critics from his Twitter account the
case turns on the question of whether the president's account at real Donald
Trump is a public forum the litigation has been a competition of analogies is
the president's notorious Twitter account like a town hall or a park or is
it more like a radio station or a telegraph Heather Whitney a philosopher
and legal theorist has written a fascinating paper about whether the
social media companies are properly thought of as editors her paper is
further evidence that we're still trying to figure out what the platforms are and
which legal labels to attach to them last year the US Supreme Court decided
an important case involving the First Amendment right of access to social
media the decision was unanimous but Justice Kennedy who wrote the courts
opinion and Justice Alito who concurred in it disagreed about whether the platforms
in their entirety should be characterized as public forums at some
level both justices seem to understand that the conceptual vocabulary available
to them was inadequate Justice Kennedy wrote quote while we now may be coming
to the realization that the cyber age is a revolution of historic proportions
we can't appreciate yet it's full dimensions and vast potential to alter
how we think express ourselves and define who we want to be he continued
the forces and directions of the Internet are so new so protein and so
far-reaching that courts must be conscious that what they say today might
be obsolete tomorrow if our collective understanding of the platforms is
limited it's a large part because the social media companies have guarded so
jealously the information that would help us
understand them over the last few years they've begun to share information about
instances in which governments compel them to remove user-generated content or
compel them to turn over user sensitive data they've begun to share information
about the limits they themselves impose on the kinds of information and the
kinds of content the users can post on their platforms just a few months ago
Facebook released the internal guidelines it uses to enforce its
community standards a release that marked with the Electronic Frontier
Foundation described as a sea change in Facebook's reporting practices these
disclosures are commendable but they're also overdue and incomplete as the
Electronic Frontier Foundation also observed perhaps under new pressure from
the public and from regulators around the world the companies can be compelled
to reveal more let's also recognize though that one reason the companies
aren't more transparent is that they themselves don't fully understand the
decisions that they're making sometimes this is by choice The Guardian
journalist who broke the Cambridge analytica story spoke to an engineer
who had once been in charge of policing third party app developers
at Facebook the engineer said he'd tried to warn Facebook about the growing black
market for its user data Facebook wasn't interested in hearing about it he said
he told the Guardian quote Facebook was in a stronger legal position if it
didn't know about the abuse that was happening they felt it was better not to
know it would be a mistake to think that
public ignorance in this context is simply a result of the company's refusal
to release information the information we most need the companies don't have
the companies are engaged in a massive social experiment with no precedence in
human history they rely increasingly on continuously
involving machine generated algorithms whose complexity defies understanding
they don't understand their own platforms and still less do they
understand their platforms broader social and political implications when
Facebook was asked about the reach of disinformation in the months preceding
the 2016 presidential election in the United States it first said that ten
million people had seen political ads posted by accounts linked to the Russian
internet research agency Jonathan Albright who's a researcher at
Columbia's journalism school then reported in the Washington Post that six
Russia link pages alone had reached 19 million people earlier this year
Facebook conceded that content posted by Russia linked accounts had reached as
many as a hundred and 26 million people if Mark Zuckerberg recent congressional
testimony showed anything it's that Facebook is a black box even to Facebook
Zuckerberg himself has sometimes acknowledged this in a recent interview
he conceded it's tough to be transparent when we don't first have a full
understanding of the state of some of our systems against that background and
now I'm finally getting to my point we should recognize that independent
journalism and research about the social media platforms is extraordinarily
valuable and that obstacles to this kind of journalism and research are
especially problematic to the extent these obstacles
impede us from understanding the forces that invisibly shape public discourse
their best thought of as obstacles to self-government we should view them I
think in much the same way we view laws that prevented us from reading federal
statutes or attending congressional hearings or accessing judicial opinions
they impede us from understanding the forces that govern us last-ditch legal
efforts to block stories before they go to print like the letter that Facebook's
lawyers sent to the Guardian aren't by any means the only obstacles worth
worrying about in this context perhaps the most
significant impediments to journalism and research about the social media
platforms arise from the company's Terms of Service which bar journalists and
researchers from using digital tools that are indispensable to the study of
the platforms at scale most significantly all the major social media
companies bar users from collecting information from their platforms through
automated means journalists and researchers can collect this information
manually but most of the platforms forbid them from collecting it using
computer code the prohibition is significant because it's impossible to
study trends and patterns and information flows without collecting
information at scale and it's practically impossible to collect
information at scale without collecting it digitally the effect of the
prohibition is to make some of the most important journalism and research
off-limits some platforms including Facebook also prohibit the use of
temporary research accounts the kinds of accounts that could allow journalists
and researchers to probe the platform's algorithms and to better study issues of
discrimination and disinformation online this prohibition too prevents
journalists and researchers from doing work we need them to do these
impediments are substantial in themselves but in the United States
they're made even more so by a federal statute called
the Computer Fraud and Abuse Act the US Justice Department understands that
statute to impose civil and criminal penalties on those who violate the
social media platforms Terms of Service on the Justice Department's reading the
law makes it a crime for a journalist or researcher to study the platforms using
basic digital tools the very existence of the law discourages journalists and
researchers from undertaking projects that are manifestly in the public
interest projects focused on for example the spread of disinformation and junk
news political polarization and unlawful discrimination in advertising anyone who
undertakes a project like that does it under the threat of legal action by the
Justice Department in the social media platforms when journalists and
researchers do undertake the projects they often modify their investigations
to avoid violating Terms of Service even if doing so makes their work less
valuable to the public in some cases the fear of liability leads them to abandon
projects altogether now I want to acknowledge right away that the social
media companies may have good reasons for generally Pro prohibiting the use of
these digital tools on their platforms Facebook's prohibition against automated
collection is presumably meant at least in part to impede the ability of
commercial competitors data aggregators and others to collect use and misuse the
data that Facebook's users post publicly Facebook's prohibition against the use
of fake accounts reflects in part an effort to ensure that users can feel
confident that other users they interact with are real people intentionally or
not though Facebook is also impeding journalists and researchers ability to
study and understand and report about the platform it's difficult to study a
digital machine like Facebook without the use of digital tools trying to study
Facebook with the without the use of these tools is like trying to study the
ocean without leaving the shore I want to return
to a point I made earlier that we should think of independent journalism and
research about the platforms as especially valuable and that we should
think of obstacles to them as especially as especially problematic a half a
century ago the US Supreme Court decided New York Times versus Sullivan which is
the landmark case that established crucial free speech protections that
American publishers rely on even today the cases some of you may know arose out
of an advertisement in the New York New York Times that solicited contributions
for the committee to defend Martin Luther King and the struggle for freedom
in the south and the ad accused certain public officials in the American South
of harassing civil rights activists and using violence to suppress peaceful
protests lb Sullivan the public safety commissioner of Montgomery Alabama filed
a libel suit against four of the clergymen who had signed the ad and also
against the New York Times which had published it the case was tried in
Alabama and the jury awarded five hundred thousand dollars in damages but
the US Supreme Court reversed in his opinion for the unanimous Court justice
brennan explained that the first amendment was intended first and
foremost to ensure the freedom of public debate on what he called public
questions quoting an earlier case he wrote the maintenance of the opportunity
for free political discussion to the end that the government may be responsive to
the will of the people is a fundamental principle of our constitutional system
and in his opinions most celebrated passage he invoked the United States
quote profound national commitment to the principle that debate on public
issues should be uninhibited robust and wide open
justice Brennan was focused in particular on speech critical of
government officials because the case before him involved exactly that kind of
speech but the key insight behind his opinion has broader implications the
insight is that the First Amendment was intended most of all to protect the
speech necessary to self-government or as Harry Calvin jr. put it and now
famous essay published only months after the New York Times case was decided the
central meaning of the First Amendment is quote to protect the speech without
which democracy cannot function now it hardly needs to be said that there is a
large distance between the question that was presented to the court in the New
York Times case and the questions I started off with today but it seems to
me that in an era in which the social media companies control the public
square in an era in which these companies are in a very real sense our
governor's journalism and research focused on these companies implicates
the very core of the First Amendment journalism and research that help us
better understand the forces that shape public discourse and that help us hold
accountable the powerful actors that control those forces is speech essential
to self-government it's what the First Amendment is for so
I'd like to offer just a few preliminary and tentative thoughts about what the
companies the courts and legislators could do to better protect the kind of
journalism and research I've been describing and more generally to ensure
that the public has the information it needs in order to understand the new
public square and to hold accountable the companies that control it if we were
committed to these goals what kinds of policy reforms might we ask for well we
might ask first that the companies be more transparent about the ways in which
they're shaping public discourse David Kaye who's the United Nations Special
Rapporteur for free expression issued a report several months ago urging the
companies to disclose quote radically more information about the nature of
their rulemaking and enforcement concerning expression on their platforms
in a report filed with the UN Human Rights Council kay
recommended that companies issue public opinions when they remove content from
their platforms so that users and others can better understand why the content is
being taken down and so that they can challenge the company's decisions when
they believe the company's decisions are unjustified those seem like good ideas
to me as Kay says quote in regulation of speech we expect some form of
accountability when authorities get it wrong we should see that happening
online too but we need the platforms to be transparent not only about what
speech they're taking down but about how they're shaping the speech they're not
taking down what kinds of speech and associations are they privileged in and
what forms does this privilege take what kinds of speech and associations are
they marginalizing and what forms is that marginalization take Kay's report
had a different focus but he highlighted the need for transparency about a
specific form of content curation quote if companies are ranking content on
social media feeds based on interactions between users here o they should explain
the data collected about such interactions and how this informs the
ranking criteria the more general point is that the companies should be more
transparent about how they're shaping public discourse they should be more
forthcoming about the forces at work in the new public square' here's a second
possible Avenue for reform we could ask the social media companies not to
enforce their Terms of Service against those who use digital tools on their
platforms in the service of bona fide journalism and research again many of
the companies themselves many of the questions that are most urgent right now
are questions the companies themselves aren't able to answer the companies
should be facilitating not obstructing the work of journalists and researchers
and others who might be able to provide answers where the companies can't even
if the companies have legitimate commercial privacy or security reasons
for generally prohibiting the use of certain digital investigative tools it
should be possible for them to create a kind of carve out a safe harbor that
expands the space for these activities while protecting the privacy of their
user and the integrity of their platforms
incidentally the data privacy regulation that went into effect in Europe over the
summer supplies a conceptual framework for exactly that kind of safe harbor as
a general matter the new regulation places significant restrictions on the
use of digital tools to collect and analyze and disseminate information
obtained from social media platforms but the regulation also encourages
individual countries to exempt journalism and research from those
general restrictions a handful of countries including the UK have already
recognized exemptions of this kind European privacy legislation in other
words distinguishes between those who use digital tools for private or
nefarious purposes and journalists who use those tools to inform the public
about matters of public concern the company's terms of service should be
similarly discerning the companies can protect their users privacy and their
platforms integrity without categorically prohibiting digital
journalism and research that is overwhelmingly in the public interest a
third possibility we could ask the courts to refuse to enforce the
company's Terms of Service against those who responsibly use digital tools in the
service of bona fide journalistic and research projects as a general matter
courts often a courts enforce contracts as written and as a general matter they
should enforce Terms of Service - but there are content contexts in which
courts declined to enforce contractual provisions that conflict
with public policy often these cases involve contractual terms whose
enforcement would disable democratic institutions or processes for example
where a contract would prohibit a person from running for office or from
criticizing public officials or from disclosing information of overriding
public importance those cases are surely relevant here for reasons I've already
explained journalism and research focused on the platform's is of special
democratic importance because of the unique role that these companies play in
shaping public discourse Terms of Service that impede that kind of journal
and research our intention with our commitment to self-government because
again they impede us from understanding the forces that profoundly shape our
interactions and our communities and our democracy here finally is a fourth
possibility we could encourage the US Congress to amend the Computer Fraud and
Abuse Act so that digital journalists and researchers can do their work
without the risk of incurring civil and criminal penalties journalists and
researchers who are investigating questions that implicate the very core
of the First Amendment's concern shouldn't have to operate under the
threat of legal sanctions let me just close by bringing us back to the
guardian story I started with that story about Cambridge analytic ah is in part a
reminder that the platforms are entirely justified to worry about the ways in
which their users data can be exploited by third parties it should also be a
reminder though of how reliant we are on the work of independent journalists and
researchers if we want to understand the platform's if we want to understand the
new public square' journalism and research about the platforms is crucial
and it deserves special solicitude under our law whatever free speech means in
the digital age it should encompass robust protections I think for this kind
of public interest journalism and research thank you
well thank you very much Jimmy offered this very
thought-provoking lecture so we have plenty of time actually for questions
now this is being very good video recorded and we have one microphone so
if you have a question please wait until I reach you so you can speak into the
microphone
we'll start us off professor Priya thank you this is really really interesting so
I have a question so you you tied the proposals that you make to the u.s.
approach to free speech and New York Times but could it be that that approach
nowadays is part of the problem rather than part of the solution so so
underlying this approach is one that proclaims those three famous words of
Brennan what is it robust uninhibited wide-open exactly but in fact you can
look at it in a different way and say it's a form of privatization so it
privatizes the regulation of speech and interestingly people who in other
contexts are very very favorable to idea that private business should be
regulated by government in this context are saying we should not regulate speech
and in fact nowadays we shouldn't even regulate those who regulate speech the
private companies that regulate speech now this may have been great in the 60s
but now leaving this domain completely on Facebook Google and so on completely
unregulated is is perhaps not sufficient to address the kind of problems that
that we have and perhaps and actually never thought about this until now the
basis for a distinction is is something like speech versus information so
so you can be as free with your speech but but these companies trade in
information and and and that's something else where government may want to be
more heavily involved with with regulating it so now of course they're
just there but so I I generally agree with you that the approach that we have
taken until now is not up to the task I don't see the problem as a problem with
York Times versus Sullivan or the you know at New York Times versus Sullivan
to the extent it just means that we want to make sure that debate about what
Brendon called public issues or public questions is uninhibited and robust and
wide open I'm all for that I don't see any problem with that I think that the
problem is a doctrinal one that is more recent and there has been a shift over
the last 50 years towards a much narrower conception of free speech which
focuses myopically on speakers right and is indifferent and you know to people
who are not in this debate it seems obvious that that free speech should
focus on speakers but it's actually not so obvious right because there are other
participants in the marketplace of ideas there are speakers but there are
listeners there also when it comes to social media platforms users their
society more generally which is supposed to be the beneficiary of this
marketplace of ideas and so I think that the the solution here is a is to reverse
that doctrinal trend and many of the most important and free speech questions
right now that are coming before the US Supreme Court at least are questions in
which there are at least arguably free speech interests on both sides of the V
right so if you think of these social media cases Facebook or Twitter they
argue that they have the First Amendment right to create the expressive
communities they want to create they are at
or they're publishing and they should be entitled to the same First Amendment
rights as any other editor or publisher and to me that's not an implausible
argument the problem is that they're not the only their First Amendment interests
aren't the only ones at stake in these you know in in these contexts they have
interest in creating the communities they want to create but users also have
free speech interests in being able to being able to participate fully in
public discourse society has an interest in ensuring that these platforms don't
undermine other important societal interests including the interest in fair
elections right and so I don't I don't like viewing these First Amendment
questions or these free speech questions through a very narrow lens focused on
speakers alone and I think that we are going to have to think about how to
reconcile or balance the rights again which don't seem to me trivial I think
Facebook does possibly have the right to create its own expressive community but
we need to balance that against other societal interests and the doctrine we
have in the United States right now it doesn't do that very well and you see
that not only in cases involving new communications platforms but you see it
in cases involved in campaign finance or cases involving commercial speech its
commercial speakers where there is this very narrow focus on the rights of the
speaker and in my view insufficient attention given to the free speech
implications of those decisions for other participants in what is sometimes
called the marketplace of ideas
Thanks I'm talking much eg I'm a visiting fellow at the school and
basically in University Belfast and I wonder whether you think there's a need
to go beyond the sort of the shield argument in much of the case you've laid
out which is a very compelling case is predicated on the technological ability
of the journalist or the researcher to work out what's going on which might
be presumptively unlawful but then trumped by one of the illustrations
you've laid out um should we also consider or expect that the platforms in
question might seek to frustrate either by not recognizing that they're
frustrating an investigation through their own data management and systems
practices or as we've sometimes seen with some of the sharing economy
companies deliberately seeking to interfere with investigation we've seen
for instance uber try to develop tools that are both clever and wicked in
recognizing when a regulator is logging on and giving them a different version
of the service and it's not wouldn't be surprising to see similar happening with
a journalistic investigation so where as just say the GDP or in Europe does give
a broad protection for researchers it doesn't do as much of some hints at us
requiring disclosure of how an algorithm operates or providing access to data as
we would for instance with classic public authorities where there may be an
access to information right or some way of requiring an explanation so does the
argument you outline lend itself to lore enough yeah maybe but maybe more of the
positive obligation school rather than the shield school yeah so this is you
know in terms of US constitutional doctrine my argument is quite ambitious
in terms of in any other terms it's a very modest argument right my argument
is ambitious only because US constitutional doctrine is so
constraining but I will not pretend that what I'm proposing is a complete
solution to the problems that I've you know described I I think that we can ask
the social media companies to be more transparent and we should
and journalism and research focused on them can create a kind of pressure to
you know a pressure that they will feel to be more you know more transparent I I
don't think we should rule out the possibility of regulation that requires
them to disclose information it's a complicated question what that
disclosure would look like when it comes to so you can distinguish between two
forms of censorship that platforms engage in I don't mean to you know I'm
using censorship in a value neutral way here the platforms decide who can be on
their platforms right the companies decide who can be on the platforms and
what can be said right by setting up the community standards so that's kind of
threshold censorship who gets who gets into the ocean in the first place right
and then there's a kind of environmental or interstitial censorship that the
platforms engage in much less visible and probably more consequential shaping
the speech by deciding which speech goes viral and which speech disappears into
the void so transparency as to the first forms of censorship is relatively
straightforward right you could demand as David Kay has and many others have
you could demand that the social media companies and you could require them to
do this disclose what speech have they taken down what speakers have they taken
down you could require them to issue public opinions explaining their
decisions although the scale presents a you know a challenge because they make
these decisions Oh thousands of times a day but that's relatively
straightforward what transparency looks like with respect to environmental or
interstitial censorship it's much more complicated and some people have called
for algorithmic transparency meaning you know the require the platforms to
disclose the algorithms that underlie those speech environments but again you
know as I said the the plan firms themselves don't fully understand
the algorithms the algorithms are continuously evolving they're based on
machine learning you know machines are adapting to a continuously changing
environment and if they were to disclose their algorithms nobody else would
understand them either and so you know you you can require them to disclose
major inputs that would be one possibility and some of the platforms
have made a move in that direction already in you know in blog posts they
will discuss how certain things end up at the top of your newsfeed and how
others get you know relegated to the bottom but that is a course form of of
transparency I think of journalism and research as output transparency you have
an input transparency right so the journalism and research isn't telling us
what the algorithms are but they're telling us it's telling us what the
results of the algorithms are and there is this many journalists and researchers
think that that form of transparency is ultimately more important because what
you don't get from input transparency even if we were show you know even if
Facebook showed us its algorithms and we fully understood its algorithms arguably
more important than the algorithms or the data that Facebook is feeding into
the algorithms and you don't get that from the algorithmic transparency the
only way you understand the interaction between the algorithm and the data is by
studying the output which is with journalists and researchers do so you
know that's just a long way of saying I think that the solution has to involve
some combination of input transparency and output transparency meaning some
combination of requiring the companies to disclose more and allowing
journalists and researchers to explore more freely and I think in the long run
this set of solutions here probably has to include a combination of
self-regulation on the part of the platforms and regulation imposed from
the outside by
by government my proposals are also modest in another sense in that I'm
focused entirely on transparency right now transparency and I maybe you know
maybe not the narrowest sense of transparency but nonetheless
transparency and but even if you had all the transparency that I've asked for
there's still the question of you know we would know what the platforms are
doing but what should we think of what the platforms are doing that's a
separate set of questions which is much more complicated than the set of
questions that I've tackled in in my view what Facebook and Twitter and
Instagram and YouTube are doing would be not worth losing any sleep over if these
companies were smaller than they are that every everything all the problems
turn on scale they stem from scale and they're that the magnitude of their
implications are societal implications as a result of their scale and if you
think if you agree with me that the problem you know the central problem is
a problem of centralization of power over the speech environment then it may
be that the most direct solution isn't a kind of free speech regulation but
rather antitrust or anti-monopoly regulation where you you know break up
the companies in some way or somehow lower the barriers to entry into the you
know marketplace in the real sense you know into this sector of social media
and their various ways you could do that you could do that again by breaking up
the companies but you could also do it through data portability requirements
you know just making it easier for people to move from Facebook to some new
platform right those forms of regulation might be a more direct way of addressing
some of the problems certainly the centralization problem and in some ways
they're less
they raise fewer difficult questions because you know regulation of speech
which is the other option would run into immediate doctrinal barriers in the
United States and I think would run into immediate value barrier it's not just in
United States but in most democracies you know people don't like the idea of
governments dictating to private entities what content you know what
viewpoints they can carry and what viewpoints they can and antitrust or
anti-monopoly options don't raise those concerns at least not to the same not to
the same extent sorry it's a very long answer - it was
probably a simple question professor Bandopadhyay I thank you to me this was
great uhm I mean I think I think your paper this is a paper does a lot already
in the sense that it moves beyond gateway regulate keeping regulation to
look at the environmental modification the sort of expansive effects of that on
both users and producers of content so but my question is that so from what I
understand as you said this is a transparency argument but it seems a
little asymmetric to mean that it's missing a dimension so for instance in
the classical understanding of free speech of the marketplace of ideas
it was ideas for their own sake and it was the more speech the better right and
you want to encourage that marketplace of ideas but and that is the classical
American understanding of it but of course we now know and as you
acknowledge that there is an economic dimension to speech the reason these
companies are powerful is because they monetize speech and they're able to do
things with speech and people listening to each other are able to monetize
information right so to the extent that your solutions if I have them right of
you know either don't have the Terms of Service or Facebook should agree not to
enforce them in case of research or court should not enforce them in cases
of research all the legislature should amend this statute to my mind they run
into not the free speech sort of the
the the old-timey free speech understanding but they run into an
economics problem which is that the term the service exists and the courts exist
and this particular legislation exists because as a matter of value people see
the economic benefits the economic side of free speech is as important and the
sort of private corporation side of free speech is as important as free speech
for its own sake and this is the kind corporations have speech to do with
campaign finance and things like that there is now an equation between these
two values of free speech for its own sake and free speech as an economic
value to society so they're both in the public interest my question is that
without measuring the economic side of this is it really to what extent would
you say it's reasonable to expect the courts or the corporation or the
legislature to want to undo or unravel the system at all
aside from there you know ideological commitment there's also an ideological
commitment on the economy side so could you say a little more about how you
think that side would play into your argument thank you that's a very hard
question so I I don't think that there is a strong argument that categorically
prohibiting these forms of journalism and research is necessary to serve
economic interests you know I think that a safe harbor for journalism and
research would add you know just to think of it in a very sort of crass
ledger you know sort of pros and cons way you know the pros would be the
public would have more information about the platforms and the cons would be very
limited even if you take economic costs into account these rules were put in
place long before anyone had thought about social media platforms a Computer
Fraud and Abuse Act is a 1980 statute the concern was hacking it's just that
the law was drafted so broadly that when the social media companies came along
they saw in the law the possibility of
you know strengthening the prohibition in their terms of service against this
form of digital journalism and research so I you know they've taken advantage of
the breadth of the law but the law doesn't reflect a congressional judgment
that banning this form of journalism is necessary or that it serves some
economic and the law was drafted before anyone you know had before Facebook was
a gleam in Mark Zuckerberg I so the hope now is that we can go back to the
legislature and say draft a narrower law you draft a narrower law that preserves
Facebook's economic viability that preserves a Facebook's ability to
protect the integrity of its platform preserves Facebook's ability to protect
its users privacy I don't like saying Facebook's protecting its users privacy
because it's ridiculous that you know but that's the argument that they make
is that they use privacy in a very stylized way what they mean is protect
our own ability to exploit our users data but prevent anyone else from
exploiting it mewho piz we can go back to Congress and say now that we have a
better sense of how now that we know how the law is being used today draft a
narrower law and with respect to so I didn't mention it but this safe harbor
idea this idea of the social media platforms creating a safe harbor
voluntarily for this kind of journalism and research you know voluntariness is
it's a spectrum right and Facebook is responding to public pressure right now
Facebook's under a lot of public pressure to respond to the major news
stories that have exposed the various pathologies on Facebook's platform that
lead to some of the things I was talking about earlier you know disinformation or
you know the undermining of elections and my Institute we sent a letter to
Facebook maybe six weeks ago proposing that they adopt a safe harbor for this
these forms of journalism and research and we spelled out in significant detail
what we meant by public interest journalistic and research projects and
we provided a draft of an amendment to their Terms of Service and I will not
say that they jumped at the opportunity to create that safe harbor harbor but
they responded graciously and invited us to come in and talk with him about it
and we've now had a couple of conversations with them most recently
just a few days ago where we spent two hours talking to Facebook's lawyers and
engineers about whether something like this would be workable and I am NOT
suggesting that you know I'm confident that this is going to happen I'm
definitely not but we would not be getting this kind of response from
Facebook if they didn't feel some public pressure now to answer the kinds of
concerns that we're that we're raising so that's just to say that we're in an
environment in which I think these kinds of reforms are at least on the table
which is more than we could have said two years ago and I think it's
worthwhile to think about how to exploit this moment especially since we don't
know how long this moment will will last
I like to address this from tools digital tools available to analyze
analyze algorithms parentsí reasons I come from small consultancy the group
that I had and I work for banks for in the corporate world government the way
they are using analysis it's a very methodical focused work for this craft
from social planners to be analyzed maybe we could borrow from their
expertise let's take an example we talk let's look at the algorithm if I don't
make sense please interrupt in here you talk input all garage you call data
coming out of the hall to the algorithm the algorithm a thematical formulas each
together I would believe as a professional wouldn't give the entire
story what will tell the story if five
Facebook would show us their use cases what does it mean a use case I want to
add a new customer I have a registration and I end up with a profile and I want
to know what goes in my profile and if you have a flag that you flagged me as a
his security I want to help you do follow-up on me yeah that's I agree with
you I so I mentioned this only briefly in my prepared remarks but so Facebook's
Terms of Service prohibit not just scraping not just collecting collecting
data through computer code they also prohibit what they call fake accounts
and in most contacts you know what counts as a fake account
is obvious you know if you create an account that's not associated with your
real name it's a fake account but researchers and journalists sometimes
want to use what Facebook would call fake accounts
and what they would call what researchers and journalists would call
temporary research accounts to probe the platform for precisely this reason so
you create an account in which you you know you you post that you're a member
of the Democratic Party and you create another account that in which you say
you're a member of the Republican Party and you can see then you know how does
Facebook respond to you you know how different are the responses and this is
a very important thing you you because the algorithms don't tell you everything
you need to know how the platform's respond to certain prompts right and
this is you know in the offline world there's a whole line of cases at least
in the u.s. that protects the ability of civil rights researchers to go to
landlords with tests you know with testers and you send one white couple to
the landlord the next day you send a black couple to the landlord and you see
how does the landlord respond differently to people who are otherwise
similar but differ in this one key quality and journalists and civil rights
researchers want to do the same thing on Facebook's platform they want to find
out well if you're black on Facebook what is the experience of being black on
Facebook and how does it differ from the experience of being white on Facebook or
you know what what is the experience of being from California on Facebook and
how does it differ from the experience of being from Toronto and on Facebook
and it's impossible to do that kind of research without violating Facebook's
Terms of Service so the safe harbor that we proposed and the safe harbor that I
referenced earlier would protect not only the ability of a researcher to
scrape the platform it would also protect the ability for researcher to
create what they would Facebook would call fake account to do that kind of
work now Facebook has said we want to create a community in which people can
feel confident that the people they're interacting with are real people and not
you know not BOTS or machines or fake fake people and so what we've proposed
is that they allow researchers and journalists certain in certain contexts
to create these temporary research accounts so long as the accounts
themselves identify themselves as fake accounts right so but there are there
are many contexts in which what you're trying to study is not the activity of
other users what you're trying to study is the activity of the platform and so
you're not really interacting with other users and to the extent you are you
would disclose to those users that this is not a real person this is the New
York Times doing a study of Facebook and you know we'll see how Facebook response
to this suggestion I mean they again they did not jump at it but nor have
they said that they're close to it good question let me now ask a question and
we have had the chance to speak about this so so your discussion is very much
couched in American medium right so Facebook is an American company you're
from an American organization and you're having a dialogue within a legal
framework that is American and that makes sense right to the extent that you
approaching Facebook on those terms and then be willing to say amend their Terms
of Service or their approach in a self-regulated way and so far as that
works fantastic the sense is that might not be the end of the story
and so the worry here in terms of strategy is having that discussion in a
context like the US which has a very specific not to say neo syncretic
understanding of freedom of speech which is as you said very user driven very Pro
corporate speech campaign-finance Hobby Lobby Citizens United cases we're all
familiar about is the worry that it the kinds of outcomes with that discussion
say whether it ends up in the court or even at the level of Congress and
tranches further some of these idiosyncrasies makes me think that it
might be worth thinking about approaching other avenues or other
venues from a transnational perspective going to the EU coming to other
jurisdictions have that discussion in order to be able
to try to have it only in any less idiosyncratic way right so we've seen
Google having trouble in the EU and my terms of services here in Canada changed
as a resolved right so so to what extent do you think it's it's it's prudent I
want to say that cab in the approach to the US and to what extent do you think
it would be worth engaging other yeah that's a good question I and I mean
you're absolutely right that you know I've approached this through a kind of
American lens I would say that the extent to which it would be productive
to approach it through a different lens depends a little bit on which reform
we're talking about right so if we're talking about pressuring the companies
to be more transparent there's no reason why that kind of campaign should be an
American campaign you know there are users of these platforms all over the
world and we're not relying on law in any direct way in pushing for this kind
of reform we are it's a softer ask and it would make sense to have a broader
coalition and that's something that we should definitely explore when it comes
to you know persuading the courts not to enforce Terms of Service it's hard to
get out of a single jurisdiction right there there the cases were a lot we'd
have to rely on in asking American courts not to enforce Terms of Service
our American cases and we're stuck with the idiosyncrasies of American law in
that context and you know there is a sense in which we could put together a
broader you know that we could have groups in multiple countries going to
their own courts and you know making the same arguments but they would have to
make the arguments in very different ways and even you know whether the
argument could be made or not might turn on what jurisdiction you're in I mean
this is a not an easy argument in the United States it might be an easier one
in countries where the public interest override to contractual agreements is
stronger you know there's more in the United States as you can imagine you
it's this very high deference to private private contractual agreements and then
this argument that you know the CFAA the Computer Fraud and Abuse Act violates
the First Amendment to the extent it prevents this kind of journalism
research there we're addressing both of them we're addressing an Americans a
prop an American problem because it's an American statute and I don't know that
there are similar things in other countries maybe there are but I don't
know and we have no choice but to use American tools to get at it there I
think the problem is especially American in that the CFAA reflects this use of
government power to protect private interests and you know that's the kind
of thing you see more in the United States than you do and your private
economic interest right but the solution our our argument that this statute
violates the First Amendment to the or violates free speech rights to the
extent that prevents this kind of journalists or research my sense is we
have more we have stronger tools at our disposal in the United States to fight
against a problem like this then you do in most other places in the world
because free speech doctrine is very well developed in the United States for
better and worse and you know it's an entirely plausible argument that this
statute could be struck down as applied to journalists and researchers on First
Amendment grounds so I think that we you know we this this set of problems would
definitely benefit from the thinking and resources and experience of a broader an
international coalition or group but the value of that might be different
depending on which reform we were talking thank you my question was about
jurisdiction too but I'm gonna know can can you take us back to the original
story and the Guardian and take us to the offline legal scene what happened
with that threat to to sue the Guardian the short answers
they back down Facebook back down I think arguments so they argued that so
the Guardian their hook was that the Guardian described this as a breach as a
data breach and Facebook the hook for Facebook's letter to the Guardian was
that it was libelous or defamatory to describe this Cambridge analytic
analytical thing is a breach because there was it's not that Facebook had
data in some you know on some hard drive somewhere and somebody hacked into it
and exposed it that's not what happened here this was a combination of data that
individual users had already made public in some sense they had posted publicly
and other information that users had voluntarily given to this researchers in
return for small sums of money and so Facebook said you absolutely cannot
publish that you know you your your story is completely wrong the framing is
all wrong this is a much more minor problem and to call it a breach is you
know not just wrong but actionable now I don't think Facebook was really
concerned about that narrow point what Facebook was concerned about was the
story you know more generally and the Guardian to its credit just published
the story and the aftermath of the story was that Facebook was you know under
fire from you know and from many different places and Facebook
predictably decided that it was probably not in its interest to be seen to be
going after the Guardian for for the story and so Facebook sort of changed
its its attitude towards the story after after it was after it was published and
but you see that you know now we represent these journalists and
researchers that are trying to study the platforms and those kinds of
conversations happen all the time behind the scenes around the Terms of Service
and the canoe fraud and abuse act like a journalist
will so we represent one journalist who
came up with a tool that allowed people to study why Facebook was recommending
certain people as friends in the friends you may know people you may know feature
so if you use faith Facebook there's a feature called people you may know and
it recommends people to you and this journalist cache hill at Gizmodo was
studying this particular feature of Facebook and interested to find out why
she certain people had been recommended to her and one of the people that
Facebook recommended to her turned out to be like a long-lost relative of hers
and she didn't even know this was a relative but she did some research and
she found out that Facebook somehow knew that this was her relative even though
she didn't and she thought well this is interesting
and so invented a browser plugin that individual users could install and track
the people who who Facebook was recommending to them and Facebook
contacted her offline and said we think that this this tool violates Terms of
Service and we think you might want to take it down and she got in in exchange
with them about you know can you explain to me how this violates the Terms of
Service because this is a browser plug-in it's not obviously in violation
of the prohibition against scraping it doesn't involve a fake account so what
what is that the term that it's prohibiting and I don't think she ever
got a satisfactory answer to that question but they tried to bully her
into abandoning this this project and she didn't you know to her credit but
many journalists abandoned projects in the face of you know the veiled threat
of legal action or not-so-veiled threat of legal action and so the same kind of
thing that you see going on with the Guardian goes on you know every day
behind the scenes and there's no no easy way to track those kinds of
conversations and you know we have no idea how many projects have been
abandoned because Facebook or some other platform said hey you know what there's
this federal statute called the Computer Fraud and Abuse Act and you know they
may as they're doing with cache Hill they may be invoking Terms of Service or
the Computer Fraud and Abuse Act even when those things don't actually apply
to the the project that the journalist or researcher is engaging so yeah that
you know that's part of the reason why we thought this safe harbor idea was
worth pursuing yeah last question hi so my name is Luke and I'm just a first
year student here at Osgoode and my question was related to the EU and you
mentioned them briefly in your speech I just wanted to know your assessment of
how they're handling it because I see that they're taking broader steps in
other countries or like North America or Canada in terms of curtailing this and
we specifically maybe - the copyright directive that they've recently done
because there's been some public backlash to that so I'm not sure how you
assess their strategy on the digital single market so I don't know a lot
about the copyright directive I think my general sense is that the EU has been
more thoughtful than the United States when it comes to privacy issues I'm
worried a little bit about their response to what they see as dangerous
speech you know so they have you know in Germany there's this new law that
requires the platforms to take down speech related to terrorism for example
within I think it's 24 hours and you know the the way these categories of
speech are defined it's often unsatisfying and the platforms have
every incentive to over censor to take down much more than might fall into
those categories because liability the liability is significant and there's no
penalty for taking down more speed than you have to and most the speech
that is taken down you know some of the speech that's taken down
you know everybody would agree this is inciting people to terrorism or inciting
people to immediate violence but a lot of it isn't like that you know a lot of
it is stuff that's controversial or involves groups that are disfavored but
doesn't involve you know sort of the kind of speech that most people would
agree governments have a legitimate interest in taking down and the cost of
that kind of over censorship often fall most on marginalized groups marginalized
because they're politically marginalized or religious minorities or racial
minorities and you know I I think that there are you know there are trade-offs
here in the United States there's a good argument to be made that Twitter for
example has which has you know used to call itself the free speech wing or the
free speech party right because they generally didn't take things down one
consequence of their not taking things down is that a lot of harassment stayed
up on the platform and there are two the groups that pay the most significant
price for that kind of harassment are women or you know minorities of one kind
or another and it'll be hard to call that a free speech success you know but
on the other hand you don't want the platforms to take down so much speech
that you know you end up with these carefully curated spaces where you know
controversial ideas are marginalized so you have to find some middle you know
some some middle middle road here on the speech issues I'm not convinced that
we've found it privacy I think that the EU seems to be
a good step ahead of the United States and on antitrust issues to the EU it's
been much more aggressive with the big technology companies and now you see for
good reasons and bad reasons the u.s. admitted the Trump administration
more interested in anti-monopoly litigation or regulation of the of the
tech companies but we're definitely behind
alright so Jimmy o started is lecture by confessing doubts as to whether his
lecture would live up to the title Omid lecture that is the light of truth now
after listening to this lecture I think you'll agree with me that he's really
made a valiant effort to shed light on an area on an issue that really warrants
more inquiry and that he's been a the lecture really champions eliciting facts
access to facts greater actor access to truth in an area where even those were
putting out the truth that his companies like Facebook generating algorithms do
not know what that truth is so it's hard for me to think of a topic for a lecture
for Nora mat lectured that was more appropriate so please join me in
thanking time will Joffrey for what was really a great
For more infomation >> City Council Business Meeting & Public Forum - November 26, 2018 - Duration: 2:13:43. 
For more infomation >> Nibiru 2018 update today | NASA Public Education - Duration: 2:59:59. 
For more infomation >> New Jersey Police Seek Public's Help Identifying Two Credit Card Thieves - Duration: 0:22. 
For more infomation >> Public hearing in Albany on student debt - Duration: 0:30. 
For more infomation >> Haslam named in 'Public Officials of the Year' by Governing Magazine - Duration: 0:25. 
For more infomation >> MPD reminds public of Internet Purchase Exchange - Duration: 0:19. 
For more infomation >> Schools save food for Public Health program fighting hunger - Duration: 0:58. 
For more infomation >> Learn about a new public health degree offered at Virginia Tech - Duration: 4:09. 
For more infomation >> Cannery Public Market Holiday Events - Duration: 3:39. 
For more infomation >> Nibiru 2018 update today | NASA Public Education 27/11/2018 - Duration: 2:59:56. 
Không có nhận xét nào:
Đăng nhận xét