Importance of Natural Resources

Section 512 Study Roundtable – Section 3


>>Okay, welcome back, everyone. I hope you had a good lunch. So, we are about
to start the third and final domestic developments
panel of this round table. The next panel will be talking about international
developments that have happened. Once again, I want to
remind everyone, you know, try to keep your
remarks succinct so that we can let
everyone have an opportunity to get their points out. If for some reason during
this panel you don’t get to say everything you want to
say, or someone says something that you feel needs
to be responded to, please do remember
there’s an open mike at the end of the afternoon. There’s a sign-up sheet. We welcome even audience members
who may not have participated on one of the panels to speak and offer their views
at that point. So, with that, we’re going to do the quick round
of introductions. Your name, your affiliation, a very brief description
of why you’re here. Why you care about this issue. And we’ll start with
Dr. Burgess.>>Richard James
Burgess: Thank you. My name is Richard
James Burgess. I’m the president of — [ Inaudible ]>>Your microphone’s off.>>Richard James Burgess: Okay. Is that on now?>>Yes.>>Richard James
Burgess: Okay, thank you. Okay, my name’s Richard
James Burgess. I’m the president and CEO of the American Association
of Independent Music. From AAIM’s perspective, things
are not that much different than they were in
2017 or before. Notice and takedown hasn’t
helped much with stay down. We have repeat infringes,
and we still have to send amazing amounts of
notices to get anything done. We’re cautiously optimistic that the EU copyright
directive might improve things, and it does establish that a
large regulatory body realizes that the spirit of the law
is not being adhered to.>>Mr. Feerst:>>Alex Feerst: Good afternoon. I’m Alex Feerst. I am head of legal at Medium. It’s a publishing platform. We have about more than
10,000 posts a day go up on Medium written by
various folks, including more than 15,000 writers who
get paid for their work. I also run a team of
content moderatos at Medium, which is called trust and safety
sometimes in the business, and I’m here, I think most of
all to help embody and talk about the moderator’s
dilemma and the way that red flag knowledge
sort of hangs over the head of my team that’s trying pretty
hard to make a thoughtful and civil space on the
internet for people to write and engage while taking down
things like violent content and terrorist content and all
sorts of other things and trying to be mindful of the
right to creators who come to our platform to write.>>Okay, professor Hartline.>>Devlin Hartline: Hi,
I’m Devlin Hartline. I’m at George Mason University. I’m here to talk about how
I think Section 512 is not intended to be solely a
notice and takedown regime. Congress intended to preserve
incentives for service providers and copyright owners to
cooperate, to detect and deal with copyright infringement. And so the idea was that service
providers would also play a role in preventing infringement, and
they would do some of the work in finding and removing
infringing content without input from copyright owners. And the main way that congress
intended for this to happen was with the red flag
knowledge standards, that is, service providers would
lose their safe harbors if they’re aware of
facts or circumstances from which infringing activity
is apparent and fail to act. But the courts have read the
red flag knowledge standards so narrowly that only a
fire engine that’s actually on fire will suffice. And one need look no
further than the fact that Google can index the
pirate bay with no concern for losing its safe harbor
to see how things have gotten out of hand, 512, on the
other hand, was supposed to be common sensical. And it wasn’t meant to
only turn on knowledge of specific infringements. Properly understood, red flag
knowledge is general knowledge. It is a subjective knowledge
of facts or circumstances that a reasonable
person would understand to be apparently infringing. And the burden of investigating
a red flag should fall squarely on the service provider. Thank you.>>Okay, thank you. Ms. Gellis.>>Catherine Gellis: Thank you. I’m Cathy Gellis. I’m a technology lawyer, and
I’ve given some oral testimony under my own auspices
at the last hearing, and I filed comments on
behalf of the Copia Institute when the comments
were solicited. There’s two rounds of that. And I’m here today
under the auspices of the Copia Institute as well. The first point I want
to make is that not to let the sky is
falling rhetoric skew the recommendations that you
may feel driven to make. The sky is not, in
fact, falling. It is, in fact, rising. The pie is growing. There is more revenue. There are more works, and all of this upward trend
is due to the internet. So we have to be really careful
not to kill this golden goose and ruin this benefit that we
have and make sure that we have in place what’s allowed
this pie to grow. The Copia Institute just today,
in partnership with CCIA, has put forth a report
aggregating a lot of the data providing
the evidentiary record for that statement. It’s available at
skyisrising.com.>>Okay.>>Catherine Gellis: But the
second point, oh, I’m sorry, I don’t want to cut
off any interest in that report whatsoever. But the second point I want to make turns the corner
a little bit and to say that even though
things are good, there’s still cause for alarm. And I’m thinking of cause
for alarm about the pressures that are falling on
platforms and ultimately on individual creators,
expressors, internet users. We’re discussing these cases, I’m imagine you’ll ask us
the questions again about BMG versus Cox and Mavericks. And in all of these cases,
there’s many differences between them and many
things that could be said. But the chief cause of alarm
is something that we flagged in our previous testimony
and also in the comments that we made, which
is, what the effect is of unadjudicated
claims of infringement. This is the only area of law
where somebody can just say, I think I’ve been wronged, and
all of a sudden, everybody needs to react without any court ever
adjudicating whether a wrong has actually happened. This is something we
need to pull back on.>>Let’s put a pin in that, and
we’ll definitely get back to it. Professor Goldman.>>Eric Goldman: Yeah, hi. I’m Eric Goldman. I’m from Santa Clara
University School of Law. I’m going to answer Mr.
Amer’s [phonetic] question about Section 512(f). I pulled all of the cases
that referenced Section 512(f) since January 1, 2017. There have been about
25 such cases. Less than half of them
actually subsequently analyzed the doctrine. Other times it was
just a stray reference. And I did not find in
my research any cases where a plaintiff under
512(f) has actually won in court during the time period. In fact, that’s not inconsistent
with past jurisprudence. I am aware of only two times in which a 512(f)
plaintiff has ever won in court in the last 21 years.>>Okay. Mr. Gratz.>>Joseph Gratz: Thank you. I’m Joe Gratz. I’m a partner at Durie
Tangri in San Francisco. I litigate cases
at the intersection of technology and copyright law. There haven’t been
that many cases since the last set
of roundtables. And I think what the
reflects that it is that the system is working,
that the players have, at least on the whole
and the main, reached a rough modus
vivendi that is leading to at least somewhat less
largescale litigation on 512 than we’ve seen in the
past in, for example, a Giganews-type case or a
Viacom-YouTube-type case. What case law there has been since 2017 has shown the
importance of the flexibility of the current law and the
ability of courts to take into account the particular
circumstances of a case, whether that’s in the context of
what it means to be expeditious and the context of what it
means under the circumstances to make a knowing
misrepresentation of 512(f) or in the context of the repeat
[inaudible] standard taking into the account the
different circumstances in, for example, Cox and Motherless.>>Okay, Mr. Polin:>>Jared Polin: Hello. My name is Jared Polin
from FroKnowsPhoto.com. I’m a content creator. I’m a photographer. As of nine years ago I
started a YouTube channel, and I create free fun
and informative content that helps photographs
and creatives around the world figure out
how to take better pictures but also make money
on the internet. I build everything off as
a free, so I give away all of this content in the hopes
that people will support me by purchasing some of the
things I do end up selling. And so, I’m here to lend
my hand as a creator, somebody who makes
a living doing this. Has two full-time employees
off of a YouTube channel that I pay benefits
for and 401K. So as a small business,
I’m here to help as the guy that’s creating
content to be a voice for, well I don’t know if reason
counts, but for something.>>Okay, Mrs. Ray.>>Tamber Ray: I’m
Tampa Ray with NTCA — [ Inaudible ]>>Go to green.>>Make sure your
light is green.>>Tampa Ray: Hi, okay,
sorry, it was on the four. I’m Tampa Ray with NTCA, the
Rural Broadband Association. Among NTCA’s members are
approximately 850 rural, small broadband providers
who cover approximately 33% of the land mass in
the United States. As the sole broadband provider
in many of their communities, they’re having a hard
time reconciling courses. The Cox and Grande
versus Peckingham cases, in many cases they have
implemented and followed through on a repeat
infringer policies. Despite these efforts, they
continue to see, in most cases, the same number, if not more,
of DMCA notices coming through. As a result, they’re
looking for guidance, looking for perhaps refinement
of the repeat infringer policies so that the purpose of
Section 512 can be carried out, and they can be aware of
their responsibilities. Thank you.>>Okay, Mr. Schwartz.>>Robert Schwartz:
I’m Bob Schwartz with the law firm
Constantine Cannon. We’re here as outside counsel to
Consumer Technology Association. In my career for CTA, we
focused on so-called gray areas of the law, and we’ve learned that the daily outcomes are
really based on toleration, accommodation and
implied license. For example, just
yesterday in my side career as a performing jazz musician, and I stress I’m
not a full-time pro like Mr. Hatfield you heard form
this morning, and I fully note and respect the difference, I fairly routinely
quoted four bars of the in copyright tune Manhattan
while heading for the end of the public domain tune
Love Will Find a Way. This is quite the
routine in jazz solos and even jazz arrangements. So the notion of, if you
curate for one thing, you need to curate, you
know, whatever is uploaded or whatever appears just cannot
possibly work in this sort of context ignoring fair
use, ignoring accommodation, ignoring all sorts of
other considerations. And if it were tried, it
would probably hurt the people that do need some help and
that it’s really intended for.>>Okay, thank you. Mr. Troncoso.>>Christian Troncoso:
Thank you, and thanks for convening
this roundtable and for undertaking the study that you guys are
in the midst of. My name is Christian Troncoso. I’m here on behalf of BSA,
the software alliance. We represent a coalition
of software companies that provide an array of
enterprise Cloud services. We bring somewhat of
a unique perspective to these conversations
because in addition to representing our
members on policy issues, we also operate a robust
online enforcement program on their behalf, because
their software is also subject to large volumes of piracy. So at the same time,
our company, in operating those
enterprise Cloud services that I mentioned
earlier are very reliant on the safe harbors themselves. So we have a bit of a dual
perspective on these issues. And from our perch, we see the
DMCA framework largely operating as intended. Certainly, there are
frustrations when you look at the framework
from one perspective. If we talked only to our content
protection folks, they have many of the same concerns that
others have mentioned today. But when we talk to
our companies in terms of their operation of enterprise
Cloud services, it’s sort of, we see the frustrations from
that side and sort of it’s one of those classic situations
where there’s a little bit of frustration on both
sides, tends to show that the compromise
was fair in the end. We think that the
DMCA, the other sort of important stakeholder in these conversations
are end users who are also subject
to the outcomes here. So, just I’ll sort
of leave it at that, and we can talk about
the case law.>>Okay, thank you. Ms. Tummarello.>>Kate Tummarello: Hi, I’m
Kate Tummarello, it’s on, policy manager at Engine. We’re a nonprofit based in D.C.
and San Francisco that advocates on behalf of startups for
pro-innovation policies. The American startup
ecosystem is thriving, in part, because of our balanced
copyright laws, including the DMCA, which
have allowed innovators to create new platforms
for users to share ideas and content. Startup support Section
512 safe harbors because they reduce the risk that a company could
face ruinous litigation over alleged infringement
occurring on its website. All issues do remain concerning
bad actors sending false notices to startup. We fully support the DMCA
framework and will continue to engage with the Copyright
Office to protect Section 512 and internet safe harbors. Thank you for the
opportunity to join the panel.>>Thank you. Mr. Wang.>>Ping Wang: Yes,
my name is Ping Wang. I’m a writer and a
freelancer for 20 years. It’s an honor to be here
today to learn from you and to share my stories
and opinions on this important issue. As you all know, the EU
has just purchased the copyright directives. I have been closely
following the development. And I have been [inaudible] the
similar US version since 2001. And I spent almost
$20,000 of my own money to [inaudible] the same thing. So I think I have a little bit
different perspective to present to you guys but I think is
necessary and important. Thank you.>>Thank you. Ms. Wolff.>>Nancy Wolff: Yes,
good afternoon. Thank you for convening
the roundtable. I’m here, well, I’m a partner
at Cowan, DeBaets, Abrahams and Sheppard, but today
I’m here on behalf of a trade association
I’ve represented for many years now
called the Digital Media Licensing Association. That’s an association
of aggregators that help content
owners individual arts, whether it’s still or motion, monetize their content
by licensing. We were here three years
ago, and I would say, I don’t think there’s been
much improvement in the balance of the DMCA when it
comes down to trying to encourage licensing
and decrease piracy. The internet would be empty
without visual content, yet the burden of the
notice and takedown exist, and the way the courts haven’t
been interpreting things such as red flag knowledge,
there’s really the lack of incentives to cooperate
and work collaboratively with the content
licensing industry, particularly in the visual arts. There’s like, there’s
much internet, there’s much fingerprinting
technology available, yet there’s little
incentive ever to do any type of filtering. And there also seems
to be little incentive to develop licensing systems. You can see from Mavericks that curation makes a
much better website, but there doesn’t seem
to be any encouragement to develop the licensing
models to make that happen.>>Okay, thank you. So since this is the
last domestic panel, we’re going to shake things up
a little bit, and we’re going to start with a topic
near and dear to Professor Goldman’s
heart, which is 512(f). And I actually want to begin
by asking Mr. Gratz a question. In the San Francisco roundtable,
you mentioned that 512(f) needed to have more teeth when dealing
with abusive notice senders. And you suggested
statutory damages. What have you thought of
the recent 512(f) decisions in the wake of Lenz, and do
you still think 512(f) needs to be changed, and if so, how?>>Joseph Gratz:
I do, thank you. I do. I think that
512(f) as the review that Professor Goldman has
recently published on his blog that I’m sure he’ll be
talking about, as well as sort of the experience
in the world is that 5129(f) is not a sufficient
deterrent for many kinds of abusive notices and
counternotifications. There haven’t been many
litigated cases on this purely because I think the,
what is at stake in any given situation
is not always enough to support litigation. But I think that we may
be seeing a greater, a greater number of those
cases coming in the future, even under the current standard,
not because of anything that comes out of Lenz,
which I think sort of restated what I think the way that most people had
already been operating. But because of increasing
and new ways that, as the internet becomes,
and as internet and immunity areas become a more
and more important part of more and more economic activity,
especially on marketplaces like Etsy and other
online marketplaces. The competitive incentives
to send bogus notifications and counternotifications
will become so large that there will all of a
sudden be an economic incentive or situations that
support economic incentives for litigation.>>Okay. Sorry. So sort of on that point, on
the last panel Etsy alluded to an increasing amount of fraud in both notices and
counternotices. And there’s been
various discussion that sometimes it’s
used to harass and for nonlegitimate purposes. And in your opening statement,
Professor Goldman, you mentioned that there have been
several cases under 512(f) since we last sort of heard
from people in this study but that no plaintiffs have won. Do you think that’s a
good thing, a bad thing? Is it a problem with 512(f)? Is it a problem with the courts? What’s going on?>>Eric Goldman: Yeah, I
think that the data suggests that either there aren’t abuses
that are worth litigating, or the laws is mis-calibrated
to protect those abuses. And I think we have plenty of
evidence to suggest the latter, that there really is
no meaningful incentive to do homework before sending
a notice, and that creates lots of opportunities for
bogus notices to be sent. And in addition,
there’s often a lot of background context
that’s taking place. One of the five [inaudible]
cases, for example, involves a dispute over a
screenplay that then morphed over into one of the marketplace
cases that Mr. Gratz mentioned. And a [inaudible] writer isn’t in a good position
to resolve that. And the 512(f) claim is really
just a chance to tweak the rival over this screenplay, not
actually to resolve the merits of the underlying
copyright dispute. It’s just an ancillary way of the parties fighting
with each other.>>And just a reminder, if
other people want to chime up, feel free to turn your card
up, and I’ll call on you. Otherwise, I’m going to
keep singling people out.>>I do have a follow-up
question to that. So Professor Goldman, you
mentioned abusive notices. And I know, relatively few
counternotices are filed. There are, obviously,
abusive counternotices too. My recollection, and I want
to be, correct me on this if I’m wrong, is the
recent 512(f) cases, I’m thinking specifically, I think it’s Johnson v. New
Destiny Community Church, are applying the same standard
to, the way their counternotices or notices, is that right?>>Eric Goldman: That’s
my understanding, yes. Right, but to your point
about counternotices as a counterbalance
against the abusive notices. There’s so many things that have
to go right for a counternotice to even be filed, let
alone be effective. First, the service writer has
to acknowledge that they want to honor counternotices. The user has to understand
those rights. The user has to feel
confident that they’re willing to submit it, and they have to
be prepared for the consequences of filing the
counternotification. So there’s a whole bunch
of things that are stacked against filing counternotices. And so to think that the current
encourages counternotices as an appropriate counterweight
against abuses and notices, I just don’t think that the
system is designed that way.>>I guess I have a question to
that, because if you’re focused on a user being competent
enough and educated enough to even follow a
proper counternotice, sort of making it
easier to be liable for filing an improper
counternotice doesn’t seem to encourage the filing
of counternotices. Or would you decouple
the two standards?>>Eric Goldman: I’m sorry, could you try that
question again?>>Yeah, so it’s the same
standard under F1 and F2, and under F-2 you’re not
seeing counternotices because users may not realize
or may be a little intimidated to file a counternotice. Will that make it harder for
you users to file counternotices if you loosen up the standard
for liability under F1 or F2 since they are the
same standard.>>Eric Goldman: To the extent that counternotices are a
nonfactor today, I don’t think that making it harder to file
counternotices will have a material impact on
the overall ecosystem. So, to me, that doesn’t really, that’s like the tail
wagging the dog. Focusing on the abuses of the
notices would be the place to start, and I think
the material impact on counternotices would
be relatively de minimis.>>Okay, Mr. Polin.>>Jared Polin: Thanks. So, on YouTube I get hit on
occasion with some claims saying that I’ve uploaded
that is copywritten, and many times I have perMs. ion to use it or it’s fair use. Then other times, I can
file a counterclaim. It’s super simple
to do on YouTube. It walks you through. It asks are you this person. Are you sure you
want to do this? Are, you know, and I
do, so I go against it. And I generally win
because we try to stick to general good practices
when we’re posting content that it’s my content,
or if we’re in fair use, acknowledging something, we’re talking directly
about what that is. And then file the counternotice,
and then it’s taken care of in a matter of days. So YouTube does a very
good job of that for us.>>Okay. Ms. Wolff.>>Nancy Wolff: Yes, I actually
think that in particular with visual content
that’s online that filing the counternotice
really puts the content owner in a disadvantage, particularly if the counternotice
is in bad faith. Because you really can’t afford
to go to court within ten days and file a claim because the
damages aren’t high enough. And there isn’t alternative
to federal court. And typically, if you don’t have
a copyright registration before that, you’d have to spend
at least minimum $800 and then find a lawyer. So I think that if
counternotices are done improperly, there’s
actually more harm to the copyright holder.>>Okay. Mr. Wang.>>Ping Wang: Yes. Regarding this issue, both
answers could be found in the question you guys raised if you keep closely
following the EU, Europe, how they developed this
copyright directive. [Inaudible] is clearly
the answer. Memes, upload filters, fair
use, legal uncertainty. I don’t want to talk more about. You can easily find
the answer online. One thing, when we talk about this [inaudible]
you mentioned YouTube. But that’s only part of it. Most, you know, companies
like Google, they earn money on this way. See from, this, [inaudible]
take advantage of this way. This is 2002. Some guys just post my
content, my writings, there. And Google just ran
it [inaudible]. 2002, 2003, 2008, 2012, 2017,
2016, ’17, ’18, ’19, last month. They keep doing that. For this case only, they
run six ads around three. It’s like a parasite
around my content. They keep, you know, it
stinks bad for this piece, it’s longer than this. I just cut it for you to review. It’s got, for this piece alone,
I’m writing mainly in Chinese, but occasionally in English. It got million clicks online. It’s a hot topic today. Only for this site. This online advertising
companies could earn tens of thousands of dollars,
but I get nothing. That’s the point. You know, you talk
about YouTube, the account they host
it on their server, oh we have some responsibility. But how about these things. I called the Google [inaudible]. Can you fix it? We can’t. They easily can,
easily, they can fix it. But why don’t you don’t do it? He smiles. Because they earn
money this way. It’s so simple. They said we can
handle the content. We can do it even on
video, letting [inaudible]. It’s very easy for them,
and I presented in 2011, many years ago, to them. I asked them, Google
production team, is this a good idea
to fix the problem. They all agreed,
the engineers agree. Why you don’t do that? They said well, we don’t know. Because it costs them money. But it’s very easy to fix. That’s one point
I want to mention.>>Can I just ask,
just to clarify, so where are those
works being posted?>>Ping Wang: You know,
[inaudible] Chinese, around the world, some
content is pasted, you know, the Chinese community in the
United States, some in Canada, Malaysia, Taiwan,
Japanese, all around.>>So different people’s
blogs or websites?>>Ping Wang: Yes, different,
for different websites. It’s just out of control. Just out of control. You know, I don’t blame
this small forums, you know. They’re just a couple
hundred people. But for Google, this big
company, they can use this, you know, they told me, they
told us, they get how much money to the content creators,
YouTube, content ID. It would be easier,
and I called them. Why you don’t do that?>>So, are you issuing, sending
a takedown notice in that case? Are you sending a takedown
notice in this case?>>Ping Wang: That’s another
issue I want to mention. I closely followed the issue. The Europe parliamentary
member, I talked to them. They told me one thing. Why should we do something to prevent you doing
something wrong? Why we have to notice you? You need to ask perMs. ion first. Without perMs. ion you just don’t
do this business. You just don’t use my stuff. Why I have to find my
stuff being infringed? Why I always have
to do something? I write, you know, hours a day. I don’t have time, see, it’s
more than hundreds of pieces and more than hundreds
of websites. I ask lawyer how much you charge
for letter, $1,000, US dollars, how much money that I have. I already paid $20,000 of
my own money to do this. I cannot keep like this anymore. Thank you.>>Thank you. Mr. Gratz.>>Joseph Gratz: Returning
briefly to the issue of why we see a relatively low
rate of counternotifications, and whether that’s consistent
or inconsistent with the idea that 512(f) needs
to be strengthened, I think they are unrelated in
that I don’t think that the, I don’t think that
anyone who ought to be sending a
counternotification is failing to do so because of a
concern about 512(f). I think that lots
of people who ought to be sending
counternotifications are doing so because of a concern that
it will potentially lead to litigation. And I mean that especially
in situations where the person is very certain that they would prevail
in any litigation. That is sending a
counternotification is sort of waking a sleeping
dragon eve if you’re really, really sure that the dragon
is wrong, because the dragon, because litigation, as
Ms. Wolff right says, in this area can b so expensive.>>Okay. Ms. Gellis.>>Catherine Gellis:
Yeah, thank you. I’m half backing
up to the 512(f) and half perhaps anticipating
that we may speak more on the Cox case later
and the 512(i) issues. The first thing I wanted to
flag is comments that we made at the last hearing
and in other comments about the First Amendment
problems with how important it
is to protect the right to anonymous speech and
that 512(g) forces a user to self-unmask, and
that is a deterrent to people unmasking themselves,
especially when they end up in cross-horns where
they’ve made an anonymous speech against a more powerful person, and that powerful person
wants to shut them up. Then the notice and
takedown becomes a weapon that that person
can use and abuse, and 512 that has no teeth, it’s
very easy for them to abuse it. And if 512(g) is the only
remedy, that’s kind of game over for that speaker. So that’s an important
point to flag. The second point is 512(f),
we’re largely dealing with, I agree with what Professor
Goldman and Mr. Gratz have said about the lack of
adequate teeth. But think also in leaping ahead
a little bit to the Cox case for the 512(i) safe harbor, we don’t even have the same
notice requirement built into the notices. So never mind that 512(c)-type
notices are not being filtered out for abuse very well
because we don’t have deterrents built in. We have even fewer deterrents
for the types of notices that end up going to
the 512(a) platforms, and then if all the
sudden that’s running into their termination policies
where they have to tally up each particular some form
of complaint that they received when there’s even fewer
deterrents to make sure that these complaints are valid. That’s a more serious concern,
especially when we look into not only that are
these notices being used to remove specific
forms of expression, but they are increasingly
getting gamed to cause a speaker
to be removed. Where people are sending
notices to platforms with the understanding
that they’re going to use up the number of
strikes that they believe that whoever spoke would have. I think there’s a lot of myth that there may be three
strikes and then you’re out. Obviously, that’s not
necessarily the law. But the more important
it is for a platform, especially perhaps the large
platforms to have some sort of specific rubric where
there’s a tally, a quota, of how many complaints
they can receive before all of a sudden they have to take
action to terminate a speaker on their site, the easier it
is for people to game that. And they’re gaming it not
just to hit a specific form, a specific item of expression,
but they’re trying to get at the whole speaker to make
sure that they get turned off from that platform entirely. And that’s something we
need to really be cognizant of not allowing to be gamed, or else it will totally
silence speakers entirely.>>So entering a point you just
made, I haven’t actually heard about this, because we heard
a lot before about, you know, YouTube competitors who issue
a largely fraudulent takedowns that their competitors’
video be taken down. By the time it gets put
back up, they’ve sort of lost their marketable window. I haven’t heard this before where you’re saying
there are people who are targeting their
competitors by trying to get them, you know, three
strikes and they’re out or whatever the strikes is. Where did you see this? Have there been studies on this?>>Catherine Gellis: I don’t
know if there’s studies. I was rummaging around
trying to find some of the anecdotes that I’ve seen. I can’t say that it’s not
competitors trying to do that. I imagine, if we have this
roadmap, this will happen, if it hasn’t happened already. Where I’ve seen it
is social media where people are saying
things that people don’t like. People may be the
type of speaker that other people don’t like. And I’ve seen righteous tweets
about I didn’t like this, so I sent a takedown
notice to Twitter. Not just because I wanted that
to come down, but I wanted to use up his quota so that
that’s a strike against him so he can’t keep doing it. And I’m remembering that
one in vague concrete terms, but I don’t think it’s the
only example I’ve seen. I’ve seen that kind of logic
percolate up in discussions where people think
that’s how it works. And I’m really concerned about
that becoming how it works. So that’s my point to make.>>Okay, thank you. Professor Hartline.>>Devlin Hartline:
So, I just want to say of course bogus takedowns
are a problem, absolutely. But I feel like the hyper focus on this is Ms. ing the
forest for the trees. So I’m looking at Google’s
transparency thing. It’s very helpful, so
they’ve been asked to takedown over four billion URLs, right. And as far as I know,
the vast, vast majority of those requests
are legitimate. So that’s the forest, and we’re
focusing on a couple trees when we’re talking about
bogus takedown notices. And in fact, it seems to
me that if the standard of 512(f) is bad faith,
I’ll bring it back to 512(f) by talking about the Lenz case. Hopefully, you’ll let me do
that, because cert was denied after your deadline, so. I just, I just want to say
that I think that the court in Lenz got it totally wrong,
and here’s why in one sentence. So the 512(f) standard
is that you have to have a knowingly materially
misrepresentation, right. So that is you have created
bad faith in your mind. The thing about good faith, what
the court was trying to bring it in under is the fact that you
haven’t formed a good faith opinion that it’s, you
know, bad material. So, but there’s a difference
between having actually have, you know, formed bad faith versus not having
formed good faith, and the court conflated
those two things.>>Okay. Mr. Feerst:>>Alex Feerst: Yeah, I
just had a couple sort of concrete experiences I
thought I’d share from this. One of them is that, so for my
team when we get a DMC notice if it’s defective or
it seems like it’s, for whatever reason it doesn’t
make sense, we’ll engage with the person who sent it
to try to help them and talk about the statue and
get the information that seems necessary
to make it compliant. I think, I’ll leave the
nuances of 512(f) to others, and I’ll just sort of say
that there’s a large number, a substantial number,
that suggests to me that, at least for some
percentage of folks, the gravity of what they’re
doing, which is sending a notice that extrajudicially
causes something to be taken down is not apparent in the
way that they interact with us, because they will say
things like, I’m not going to do your job for you and
send an incomplete notice without the URL of the
infringing material. Which is an extreme, but,
and relatively unusual. But I think that the tenor of
all of this is I think for us, we want to try to help
folks give us information that will cause us to take
down anything that’s going to be inadvertently on a site. But the relative sloppiness
or sometimes the lack of complete information that
we get is just the frustration that we try to engage
with people on, try to get all that information. But I think –>>Can I ask you
[inaudible] question.>>Alex Feerst: Yeah.>>So it sounds like you
feel that the provision of the law giving you some time to expeditiously remove material
is forgiving enough to allow you to have this back
and forth with sort of potentially wonkey [phonetic]
takedown where you can figure out if it’s really
a copyright claim or if it’s really a
takedown that you’re going to honor before, as
opposed to an obligation to like act immediately.>>Alex Feerst: So,
I’m not sure. I mean I think I would say,
the way that you’re framing it, that is risk that we
would have to take on because the definition of
expeditiously is something that might be defined
later in whatever form. So I think in the instances
where we take on additional risk in order to help folks, I
think that’s what we’re doing. Whether that’s proper affair
I sort of leave to others, but I think you’re saying
that we have the time to do it is really the fact that
we’ve made the determination that we’re going to take on the
risk that we might be deemed to have been taking too long.>>Yeah, that you’re
comfortable assuming the risk that these actions you’re
talking about all would then, you know, still qualify
as expeditious removal.>>Alex Feerst: I’d say
it’s because we care. I don’t know if we
do it comfortably. But it’s something
that we’re doing.>>Okay.>>Okay, I’m going to let Dr.
Burgess as well, but after that, I’m going to switch to a
slightly different portion of 512(f) and talk about
automated notices versus forms versus other ways of doing it and whether there’s
been any change. So, Dr. Burgess.>>Richard James Burgess: Well,
I wanted to express concern about the emphasis
on hearsay and rumor that I’m hearing coming from Ms.
Gellis and also just something that she brought up earlier. I mean we need to go back
to first principles here. I represent more than
600 independent labels. The majority of those
labels have given up on the notice
and takedown system. It’s not working for them. They can’t afford it. And these are not
sole proprietors. They’re reasonably
successful independent labels in the most case, but
they’ve just given up, and –>>Do you just agree
with Mr. Gratz that the last couple yeas
have shown a modus vivendi of how we can get on
from the perspective of the individual labels?>>Richard James Burgess: I’m
sorry, I couldn’t catch that.>>Like Mr. Gratz said starting
out, you know, the last couple of years there’s been relatively
few big cases, and that kind of shows everyone has
figured out how to work within the system
voluntarily, to –>>Richard James Burgess: No, it’s that people have
given up on our side. And they’ve just, that’s
why I say, you know, we sort of have some hope that maybe the new copyright
directive might actually shift the conversation. But then Ms. Gellis said
something in her opening speech which is that the internet
is the golden goose and that, you know, the industry
is rising. And actually, content is the
golden goose, not the internet. And it’s worth remembering that,
you know, the music industry, I’ll just speak about the
music industry, for example, is not on a retail
basis somewhere around the 8 billion mark. In 1999, it was unadjusted
dollars about 21 billion. It was 14 billion at the
time, about 21 billion today. So the music industry is still
only slightly above a third of what it was 20 years
ago before it was disrupted by pirates. And we still have that
situation, basically. So, it is fine to focus on these
counternotices and everything, but when I’m hearing
immense amounts of rumor and hearsay and, you know,
voices that are obviously funded by these platforms that benefit
from the piracy of our content, then I just wanted to go
back to first principles. We own this stuff. We should not have
to be responsible for other people’s
wrong use of it.>>Okay. Mr. Polin, I’m
going to let you respond, but then we are switching
topics.>>Jared Polin: Thank you. I just, as a content creator who
gives away a lot of my stuff, the one thing I don’t want to see happen is the gatekeepers
shutting the gate again that we’ve been able to open
up through the platforms that we have access to today. That gives me the
ability to make money off of free and then selling. We talked a lot about
musicians not making as much money as they used to. Well a lot of them, some
of them don’t deserve it. They’re not working hard
enough some of them. You know? They’re not
doing what they need to do. They’re stuck in the old ways. And so you’ve got to use
what’s here today to grow. I work with a lot of musicians, and a lot of the
musicians I photograph, I have a larger online
presence than they do. And I built it off of free. So what I don’t want to
see is a gatekeeper come and shut the gates again
on us for being able to do what we’re doing,
and that’s, you know, content is a major thing. So, that’s it. That’s what I got.>>Okay. And so, sort of
switching gears a little bit. There was a lot of discussion
in the prior roundtables, and I think no consensus,
about what is the optimal way to have a notice
and takedown system. On the one hand, we hear people
like M. Feerst complaining that some of the notices
they get are sloppy or are not complete. We heard some discussion that sometimes that’s maybe
more smaller content creators than larger ones. We heard other ISPs
complaining about the volume of automated notices
they were getting from some of the larger ISPs. And then we had yet more
individual content creators complaining about, for example,
Google creating a form rather than allowing notices
to come in via email. And I think Ms. Wolff,
DEMA [phonetic] mentioned in their initial comments
that they had some concerns after Lenz as to whether that
was going to shut the door on automated notices
or whether it was going to somehow bring the
system grinding to a halt. And so I’d like to hear
people’s experience, has anything changed? Is there more consensus
about what the best way to handle these notices are? Are there preferences? Is it hopelessly lost? Mrs. Ray.>>Tamber Ray: Thank you. I’d just like to speak to the
issue of the volume of notices and where they’re coming
from and to the notion of whether streamlining
would help. Many of our members have
commented that the volumes in some cases come repeatedly
from different email addresses, even though they may have and do have a single DMCA
registered email address, and that’s where the notices
are supposed to come to. They’ll come to there. They’ll come to multiple
email addresses. And they may or may not
be identical DMCA notices. So, they are having to
struggle to compare the notices, first of all, to make sure
they are, in fact, duplicates. If they’re not, of course,
then they act on them. The other difficulty
that they’re running into is many times they will get
notices for the same subscriber for the same song from the
agency representing the copyright owner. Resulting in multiple
notices for the same song for the same subscriber. So to the extent that those
can somehow be minimized to more effectively identify
where the concerns are, where the copyright
infringement may be occurring, because from an ISP standpoint,
looking at the court cases that have come around, and
similar to what we were talking about earlier in the,
you know, 512(f), is what about these instances
where you get the same notice, adult multiple email addresses, but they don’t count
as just one notice. So, to the ISP and to the court and to somebody challenging
the ISP in court, they’re going to say hey, you’ve got
hundreds of notices for this single subscriber
in one day, and you only acted on one. While they could probably,
perhaps, hopefully, defend that, it’s going to take a whole lot
of recordkeeping on the part of the ISP and a whole
lot attorney’s fees.>>Okay. Ms. Wolff.>>Nancy Wolff: I
would say, yeah, that I’m not sure Lenz
has changed the landscape for members of DMLA. They may use image recognition
technology to find matches, but there’s always been a level
of human involvement to review. I mean, it’s very
complicated, not easy to check because you have to check
whether a work was licensed or not. And I think there
always was a look at whether something could
legitimately be a fair use or not. I mean I have recommended
that members actually assert that in their letters to say
they have considered fair use. I think the real problem is
just the entire system itself, and I agree with Dr. Burgess that I think many have
just given up on doing it. Because there’s only so many
hours in a day, and the members of the organization represent
content owners that are in the business of
trying to make money, and they could spend all
day searching the internet and trying to determine,
you know, who has licensed
what and who hasn’t. But I think they feel very
frustrated at the process. Not just 512, but you know,
the expense of litigation and, of course, have been
very much in favor of having a copyright
small claims for the infringement areas.>>Okay.>>If I could jump in here
with a follow-up to that. Now the second person,
very different area of content that’s mentioned this
frustration and maybe dropout from the notice sending
a system. In 2016, Google’s
representative in San Francisco, Fred von Lohmann, mentioned
that the high volume of notices shouldn’t
be taken as a negative but actually a sign
the system’s working. Since then, which is
right about the time that Google’s notice reception
peaked at about 20 million, I think it was per
week or per month, 20 million takedown
requests per week. Now, it’s down closer to
I think about 15 million. And I was going to try to
put on the spot Mr. Gratz, who one of his hats obviously
is as a Google lawyer. Could you give any, shed any
light on what you make of that?>>Joseph Gratz: Thank
you, I want to clarify that I’m here on my own behalf. As you say, Mr. Greenberg, I have previously represented
Google in litigation. So, that, I think that
with respect to the number of takedowns sent to Google
or the number of takedowns that are reported in
Google’s transparency report, that number verges on
meaningless because takedowns that are, the vast bulk of those
takedowns are for sites that, or items that were never indexed
by Google in the first place or never appeared in
Google search results. That is, people are using the
Google search result takedown form as a general, you know,
I haven’t seen this anywhere. But there is infringing material
somewhere on the internet. And so, to that extent,
I don’t think that the number reported there
as a bell whether of anything in particular is likely to
be particularly meaningful, particularly at the difference
between double digit millions that are probably mostly sent
by a relatively small number of automated [inaudible].>>Okay. So we’re
going to go with one of our quiet folks,
Mr. Troncoso.>>Christian Troncoso? Yeah, I just wanted to briefly
return to your original question about sort of what would
be the ideal structure for a takedown framework
in order to create the most sort
of efficient system. I think part of the difficulty
in answering that type of question is the sheer
diversity of stakeholders that have to operate
under the statute. And that’s true, but from
the service provider side, you have search engines, you
have social media companies, then you also have
enterprise Cloud companies who are 512(c) companies
who provide services to 512(c) companies that
sit on their infrastructure. So coming up with a single
system that is going to work for all of those stakeholders
is very difficult to imagine. At the same time, on the content
side, right, you have a lot of different types of
content within industries that have different
licensing process and how they like to interact with services. So coming up with one ideal
framework probably isn’t going to work, which is why we think that DMCA has done a fairly good
job of allowing for bespoke sort of systems to arise,
so you have, Mr. Gratz mentioned earlier,
Google is doing things that are not necessarily
required by the statute like de-indexing content
before it has even appeared on their services. The DMCA provides the incentives
for those types of behaviors because Google has the
incentive to make sure that they are limiting the sort
of resources that they need to throw at a problem
by dealing with it sort on the front end
before these sort of links have populated
on the service. So, again, just making the point
that sort of, there’s a lot of actors and sort of
balancing the interests of all of them is not easy, but we
think all thing considered, that DMCA has done
a pretty good job.>>Do you want to? Okay. I see we have one more
person who hasn’t spoken yet who wants to
speak, so Mr. Schwartz.>>Robert Schwartz:
Well, Google is also, of course, a member of CTA. And CTA’s view has been that
with all of these variables and difficulties, the burden
has to be, for initiation, has to be on the rights holder. Now in a previous round, CTA made a public comment
discussing the progress that has been made, not only in
automated filtering, you know, based on agreements and material
provided by content owners and Google’s technology, but also in automated
takedown notices and how all of that was working. And in discussing
progress in this area, one of the organizations of
content holders jumped in and said, well, its glad
that CTA agrees with us that the burden, they’re
doing so well at this, the burden really needs to be
on the host and the platform, not on the rights holder. No. The top of the funnel,
knowing what’s there, knowing what the
complexities are, and it’s tough for everybody, has to originate
with the rights holder. I’m sure I’m not going to
persuade everybody on this panel of that, but I’m here for CTA, and that certainly has been
CTA’s position consistently.>>Okay, I think the order
request went Professor Goldman, Mr. Wang and then
Professor Hartline. So we’re going to go with that.>>Eric Goldman: My apologies. I’m going to state the
obvious that there’s a lot of activity that’s now
outside the scope of 512(c) and 512(f) altogether. And so it’s a little hard to number crunch based
on historical data. And it also is a reminder
about the waning scope of both of those laws because
of the fact that there are two
different ways in which actions taking place
that isn’t regulated by them. First is what I’ll call the
fast lane for rights holders. That services may
develop systems that allows rights holders to have more trusted
interactions with the service. And when they interact with the
service using the tools provided by the service, then their
actions are not governed by 512(c) or 512(f). The alternative thing
that’s taking place was as mentioned these
pre-filtering type removals that never get live to the site. Those aren’t taken down
by 512(c)(3) notice and not regulated by 512(f),
and so there’s a lot of activity that is outside now the scope of
this entire regulatory scheme. So to your answer about can
we develop an optimal scheme for 512(c) or 512(f)? We have to acknowledge how much
activity is taking place outside of the scope of those, and then
we can draw our own conclusions whether that’s actually a good
thing that the DMCA facilitated or a problem because now all
this activity no longer is being properly regulated by
Congress and scheme.>>Okay, Mr. Wang.>>Ping Wang: Yes. As to see the transparency
and the taking policy. Google always, and YouTube, always claim how much money
they gave to content creators, how many links they removed,
how many videos they removed. But they never tell us
how much money they earned by using content creator’s
work without perMs. ion. I think it’s
technically doable. If everyone says oh,
this is too complicated. Let’s go one step by
step, one step by step. How much money you earned
each year or each quarter by using content creator’s
work without perMs. ion and the links, the
videos you removed. Oh you removed media,
you keep the money. That’s the best business
I have ever heard. I just take your PC, go out for
a while, then I take it back. That doesn’t make any sense. My seven-year-old son
took a book from the gym, no one’s there, and he’s
reading and he returned back to more money first thing. He said, I don’t
want to do this, Dad. Then later today, [inaudible]. Even seven-year-old
boy understand. This is not the right
thing to do. You need to ask perMs. ion first. Okay. I think Google should
do way, way more than that, just taking down and
then keep the money. At least I have the right to know how much money
you earned using my stuff without my perMs. ion. Did I ask too much? You can keep the money, but
just tell me how much money you earned. Is this a fair question?>>Okay, thank you. Professor Hartline.>>Devlin Hartline: All
right, three quick points. Lenz is wrong. The absence of good faith is
not the same thing as bad faith. So it’s wrong on the merits. But it’s also wrong as a
matter of policy, right. And it makes no sense if we
have to consider fair use, then why don’t we
have to consider all of the other possible defenses, and the court doesn’t
have an answer for that. But even within the
Lenz framework, I think it’s totally
possible to use an algorithm to assess whether you can take
something down in good faith. And so I believe that with
an algorithm you can tell that something is infringing,
but it’s much more difficult to tell if something
is not infringing, i.e., if it’s fair use. So, for example, the
entire Harry Potter movie is on a website. You can, the algorithm
can take that down. That’s good faith, within Lenz. The second point is –>>Can I, before you go
to your second point.>>Devlin Hartline: Sure.>>So, just to make
sure I’m clear. So are you arguing that Lenz
is wrong both as a matter of legal interpretation
but also policy?>>Devlin Hartline: Correct.>>Okay, now on the first
point, I mean, could you sort of walk me through why you
reached that conclusion? I mean if the statute
says that you can’t, that in 512(c) there
has to be a statement that there’s a good
faith belief that the use of material is not
authorized by the law –>>Devlin Hartline:
Yeah, and so –>>So why is making
that affirmation not, without having made an inquiry, not in knowing a
misrepresentation?>>Devlin Hartline:
Because the absence of good faith isn’t
the same thing as having actually
subjective bad faith, right. So the 512(f) standard
is subjective bad faith. The 512(c) standard is
you have good faith. There’s a difference.>>The 512(f) standard is a
knowing misrepresentation.>>Devlin Hartline: Right.>>If you know that you
have not acquired –>>Devlin Hartline: It could be, but it’s not always
[inaudible] is what I’m saying. And so under the 9th Circuit’s
reasoning, they were equated. There can be overlap.>>Anyone:>>Devlin Hartline: It’s a distinction a
lawyer [inaudible], right. I’m sure many disagree.>>Anyone else want
to weigh in on Lenz? Mr. Gratz.>>Joseph Gratz: Very briefly. A good heart and an empty head
is not good enough for 512.>>Okay. Now I’m going to let
Dr. Burgess speak to this topic, and then we’re going to move
away from 512(f) into some of the more neglected
sections of 512. So, Dr. Burgess.>>Richard James Burgess: Well,
I just wanted to say the idea that DMCA is working
is not supported by any content creators
that I know. And by the way, and
I’m an artist myself. You know, I’ve signed to six
major labels in my career. I still see royalty statements. So I have very firsthand
knowledge of how this is all working. And to Mr. Polin’s
point about gatekeepers, I mean Google is
the new gatekeeper. I know this roundtable is
loaded with Google advocates. But, you know, Google,
for example, YouTube threatens
our label as not to promote their
content unless they agree to all their content being
available on their free service. DMCA was designed to work in
a completely different world. UGC did not exist when
the DMCA was written. And the idea that it works,
I mean, it doesn’t work for content creators at all. It’s not working right now.>>Okay, thank you. So moving away from 512(f),
I want to start talking about two sections of the
DMCA that we didn’t hear a lot about at the initial
roundtables, and we haven’t heard
much about today. And those are 512(i) and 512(j). And so starting with i, during our time period
there was one case, Strike Three Holdings
versus Doe. The court found a good
cause to issue a subpoena because the plaintiff did
not request a subpoena from the clerk of the court
pursuant to 512(h), but instead, filed a lawsuit that seeks the
John Doe defendant’s identity through 45 subpoena. Does this standard run contrary
to Congress’ goal with the DMCA to encourage cooperation between
the online service providers and the rights holders
and to reduce the burden of pursuing remedies in court? Is it correct? Is it wrong? Was the court misapplying
how they should be thinking about subpoenas under the DMCA? Who wants to take that? Ms. Gellis.>>Catherine Gellis: Well,
so, there’s another provision which we’ve always somewhat
ignored which is the 512(h). And I know that at the last
roundtable hearing I complained about it. Mostly because I don’t know
if there’s actually been a ton of case law that’s really
percolated up certainly not to an appellant level,
and you tend to see things that are happening more at
the beginnings of litigation. So things wash out
at that point. But I’ve been very, I
think it’s really important to drive home the fact
that what we are talking about with the internet
is expression online, and that expression
is protected, and that includes the full
suite of First Amendment, expressive rights, including
the right to anonymity. Even outside the copyright
context where you deal with people trying to subpoena
the identity of a speaker, there are some significant due
process problems that arrive when we start to
litigate through that. But there seems to be
even graver problems when we start putting it
in the copyright context of what the due process
protections are. In terms of the Strike
Three Holdings, I did look at the case before I
came here, and I kind of threw up my hands in horror
because none of it, whether we’re leaning
towards Rule 45 or whether we’re leaning towards
512(h), I’m not convinced that there’s adequate
protections built in for users, and how ever we consider
the subpoena context in copyright cases, especially
arising out of the DMCA, we need to make that a point of
emphasis to protect that right to anonymous speech
and make it meaningful from a due process perspective.>>Okay, and just to
clarify the record, since I saw some very
confused looks over there. It is 512(h) that
we’re talking about. I was reading, I was
reading from my outline, and it’s numbered i, and so
I just flipped them around.>>Strike Three Holding cases.>>So no, we are
talking about 512(h). Does that mean anyone
else has comments that they would like
to, Mr. Gratz.>>Joseph Gratz: One brief
comment with respect to 512(h), there have certainly been a
large number of 512(h) subpoenas from Strike Three Holdings. One thing that on,
in the context not of that particular case where
the subpoena was denied, but in the context of
a number of other cases where subpoenas have been
issued, one thing I think to note about, in
the, a development in the 512(h) sphere is
that many courts have, because of the potentially
embarrassing nature of the copyrighted
material at issue in the Strike Three Holdings
cases, have placed protections in place to allow copyright
infringement cases to proceed without permitting the public
disclosure of the identity of the accused infringer. And I think that’s an important
development that helps separate out copyright interests
from anonymity interests and prevents copyright
cases from being brought, not for a legitimate
purpose of seeking recompense for infringement but in order
to hold up someone’s reputation in order to disclose that
they like a particular kind of copyrighted content. The other thing I want to
mention briefly, and I want to, I know she is on the next
panel, but Ms. McSherry of the Electronic Frontier
Foundation is counsel in a case where the recipient of a 512(h)
subpoena is moving to quash that subpoena on the
ground that the very, the very thin [inaudible]
interested issue there, that is posting a portion
of a published work in order to criticize the
public work itself, doesn’t counterbalance the
anonymity interest there of someone criticizing
a religious group that is seeking the
identity of the person who may still be a member
of the religious group and face serious social
sanction rather than any sort of copyright-related concern.>>Okay, I think I saw Ms. Wolff
and then Professor Goldman.>>Nancy Wolff: Okay. Three quick points. One, unfortunate that too
many copyright cases come out of the pornography industry. Sometimes I think it
skews some decisions. I think just uploading
content should not, per se, be equivalent to making speech. You know, if you’re commenting or criticizing, that’s
one thing. But I think just, you know, uploading other people’s
content is not the equal, I think, of speech. The others, I think, the ability
to make subpoenas is going to be more and more important
because particularly, you can’t find identities
of the, of hosts and sites to even do takedown
notices sometimes, or to contact infringers because
of the privacy laws coming from Europe that many identities
are now, you know, protected. And you wouldn’t even
be able to determine, even if you’re outside 512 who,
you know, who’s using content.>>Okay. Professor Goldman.>>Eric Goldman:
Yeah, thank you. And I appreciate your
clarification, because I was one of those puzzled people –>>I saw that.>>Eric Goldman:
Because I thought, I know the 512(h)
Holding on Strike Three. So just a few points, 512(h) is
an artifact of a different time. If we were to draft
a statute today, almost certainly we would
not draft it the way it’s currently drafted. And part of that’s because we
really don’t provide the same kind of court unsupervised
access to identity in other types of
legal doctrines. This is a very
copyright-specific solution. In almost all other
circumstances, a person who wants to identify
a tortious actor must go and get the court’s perMs. ion. But 512(h) says you
don’t need the judge’s perMs. ion at all. You automatically get
as a matte of right. I don’t think that’s a deal
we would strike today, and so, 512(h) is, I think,
just an anachronism more than anything else. I’d say that 512(h)
is fasling [phonetic] to get an identity has
become one of the sources of copyright trolling. That it has enabled people
to go and bring lawsuits with this sole intent of
extracting settlements, not of litigating them. And that’s why the particular
ruling you mentioned was so [inaudible] because the judge
said I’m not going to do it. I’m not going to allow
you to have that kind of unfettered access to identity
knowing that the consequence is that you’re not going
to come back in court. You’re going to resolve
this extra judiciously without my supervision, without
necessary legal merits to do so. I will also point out that
privacy law would make 512(h) look even more of anachronism. We’ve seen that in the who is
context that the availability to sue people who might be
alleged trademark infringers for domain names
is now not going to be automatically
given either. And in fact, most domain name
registrants now use a privacy protected service anyway to
overcome the who is model. Those were the models
of the 1990s, and 2010s would have much more
privacy protective approach in 512(h).>>So speaking of
potential anachronisms, I mentioned earlier 512(j). That’s one section of
the statute that sort of nobody talks about. There are very few cases. Is that just DOA now? Is there anything left? Do we care about 512(j)? Should we care about 512(j)? Professor Goldman.>>Eric Goldman: I have logged
almost every 512(j) case I’ve seen, because I do find
them so interesting. Because 512(j) to me
was an integral part of the DMCA structure. The DMCA safe harbor
does not say that service writers
can avoid liability. DMCA safe harbor says that they
can avoid financial damages and be subject to a
limited injunction which is spelled out in 512(j). So for me, the idea is that
512(j) is available to all of the copyright owners who are
upset about infringement online. There’s still the possibility
of exercising the rights that are permitted under that. And I do not understand why that has not been
more widely explored. That does not make sense to me.>>Okay. Mrs. Ray
and then Mr. Gratz.>>Tamber Ray: Thank you. And my comment just goes back
to the previous one on 512(h). Just wanted to echo their
comments and their support. The notion that with the
subpoenas what we have found a lot is that these subpoenas
are used, unfortunately, instead of to protect
copyrights, they ae used to gain access to personal
information that is then used to try to reach a
settlement, try to embarrass. And we are just trying to find
a way to get support for that as well as in the instances
where some of these cases where they’re actually asking
for the information it pertains to individuals who aren’t even
in that particular location. That they are outside of
the court’s jurisdiction. So, you know, while the goal
of the act was, you know, to try to provide another
avenue for protecting copyrights and give them another,
a method of enforcing, a method of resolving it, most
cases we’re finding the same as they had, that it’s not
being used for the intended way.>>Mr. Gratz.>>Joseph Gratz: I have been
puzzled about the same thing, that is, why is nobody
using 512(j)? My best answer I’ve been
able to come up with is this, the lack of cases in which
512(j) is invoked is an indication of how
cooperation has happened between online service
providers and centers of online service providers and
centers of takedown notices. In not in the sense of
necessarily holding hands, singing Kumbaya, and
trying to do things that are the best possible
thing that could be done for the enforcement
of copyrights. But instead, in that, what 512(j) gets you is you
can force the online service provider to do the thing that the online service
provider is otherwise provided with pretty good
incentives to do, and we want the online
service provider to do. That is, to take things down,
to terminate subscribers who are involved, who are
habitual infringers and so on. And because those things
happen anyway without the need of a 512(j) injunction, we don’t
see a lot of 512(j) injunctions because the purpose of
that injunction has already been achieved.>>Okay. So we’re coming down
to our last 15 minutes or so. And I know that we’ve had
a very lively discussion on earlier panels about repeat
infringers, so I’m going to give all of you
the opportunity to discuss repeat
infringer as well. Obviously, you know, we’ve heard
a lot about Cox, Motherless, Grande, how the three of
them can be reconciled. Can they be reconciled? Did they get it right? Did they get it wrong? I’d be interested to hear any
panelist’s thoughts on that.>>Ping Want: Getting
along what? You mean the providers and
the content owners, right?>>It’s on the repeat
infringer policy. So requiring that ISPs take
notices in of infringers and then ultimately
terminate their accounts. Ms. Tummarello, you
haven’t spoken today.>>Kate Tummarello: Yeah, so I think an earlier panelist
described it as a spectrum between Motherless and
Cox, and we appreciate that flexibility considering
startups often don’t have the manpower to kind of deal with this the way a
large company does, and I think earlier someone said
Motherless was run by one person who was dealing with all this. So definitely appreciate
that flexibility. That being said, I think
a necessary condition for startup growth is
certainty and clarity on repeat infringer policy
would really help companies to make sure they’re
operating within the lines. And when they’re trying
to do the right thing, know that they’re
doing the right thing and can tell investors that
they’re doing the right things. That they don’t have to worry
about litigation down the road.>>So what would that type
of certainty look like? Are you talking about
changes to the statute? Are you talking about
guiding documents that the industry comes up with? I mean what is?>>Kate Tummarello: I
think there’s a bunch of different avenues
that could work here. And I don’t want to say one is
specifically the right thing. Obviously, anything that could
be, you know, consensus-based and incorporate the perspective of small companies would
be really important in developing something that like an industry
guidance document. But we’d want to make sure
that there’s flexibility here, and there’s a chance for
startups to have input so their perspective is seen.>>Okay, Mr. Schwartz?>>Robert Schwartz: Well,
for CTA and also CCIA, we got into the Cox
case as an amicus. It was on the contributory
infringement side. Others addressed the 512 side. But in, generally in getting
into that case and looking at what happened, one thing that hasn’t been mentioned is
the terrific abuse that was part of the notices that
Cox received. They out of hand
rejected the ones that had license offers attached
to them and other things, which I think the courts
have generally taken a dim, a dim view of. That’s not to justify, you know,
a lot of the unfortunate stuff that came into the record. And the other thing we
observed is that when that unfortunate stuff of,
you know, people saying ah, the heck with it, or we’ll
reinstate the guy the next day, came into the record when the
judge said 512 has nothing to do with this case when they got to the contributory
infringement trial, it all came in as metrics
having to do with 512 compliance and have nothing at all to do
with contributory infringement. And you can search the record
for anything that really did. But the jury, nevertheless, was
allowed to consider all of it. So, in view of the
shortness of time, that’s really what
I can add on it. But I think it’s a couple of
points that have stuck with me since we got involved
in that case. And it wasn’t completely cleaned up on the contributory
infringement side by the fourth circuit.>>Okay, Ms. Gellis,
then Mr. Gratz, then Mrs. Ray, then Mr. Feerst.>>Catherine Gellis: Thank you. Well this is what I was alluding
to in my opening comments. And the Cox court, the 4th
Circuit, dealt with this head on where they basically, Cox’s
argument was that accruals for purposes of termination
policies should not be contingent upon anything
but adjudicated claims. And this is the problem with
the DMCA system to the extent that an alleged claim can
cause such sanction and effect for the platform, for
the user, for expression, for a speaker’s future
expressive rights. And this is a problem. The Cox court didn’t
seem to think that this was a big problem,
but this is a big problem. Because this essentially
becomes prior restraint where you are causing censorship
on a non-adjudicated claim. And I think that’s not going
to pass First Amendment muster. The second point that’s related
that is also an uncurrent through all the cases we’ve been
discussing today is how it’s not black and white. Sometimes maybe infringement
is maybe black and white where it’s clearly
infringing or clearly not, but we’re not certain. And I know in the first session
we were discussing even whether you could derive certainty
if you saw a watermark. One of the differences between
the Maverick’s case and the case that came afterwards was
the second one didn’t have watermarks, the first one did. And the court was
critical in the second case of well why did you even
put a watermark on it? But you still can’t
get that certainty. And any sort of, when you’re
looking at a bit of expression, the things that you
need to figure out whether there’s infringement
is is it copyrightable? Who owns it? Is it jointly owned? Who posted it under license? And was it posted as a fair use? And this is, and there
is other considerations. And the senders are
getting this wrong. If the senders are
getting it wrong, the platforms won’t have any
hope of getting it right either. But they’re being forced
to essentially defer to the complain and
presume it legitimate and take actions accordingly. And those actions are hostile
to individual expression, and especially when
we start talking about termination policies to the future expressive
rights of users. Just to double back to
what I was mentioning with that Twitter anecdote. I’ve also seen it in the
contact with people on YouTube who are worried not just that
their video got taken down, but it was a strike accrued. I’m sorry I don’t have
direct data on this, but I think they’re, I think
there are accruing anecdotes of people stressed
out that it’s not just that they’re losing the
individual expression but that they’re
losing the ability to make future expression,
because they’re stuck in a system where there’s
been some injustice where they think
their video was fine, and the problem is not just
fine, that video is down, whatever, but that it’s
going to affect their ability to make future videos and contribute future
expression to it.>>I’m sorry, go ahead.>>I mean, when you’re
Reading Cox, you’re not really like struck necessarily
by the expression of these repeat infringers. And we’ve also heard earlier like don’t make the
perfect the enemy, the good. I mean, do you have any
room for there being like a continuum here
for the [inaudible] to employ these policies?>>Catherine Gellis: Well one of the things I believe Mr.
Schwartz was just pointing out was in Cox itself,
there were severe problems with significant numbers of
the notices that were received. In fact, the court itself, the district court dismissed all
the claims based on, I believe, it was the Round
Hill Publishers, as not being valid
expressions of copyright claims. It took an enormous amount
of, in the overall record, there have been voluminous
notices sent to Cox. And the court itself crossed
out a considerable volume of those voluminous notices as not being valid
expressions of copyright. And then ignored what the impact
would be because it acknowledged that there were problems with so
many of them but still managed to default to the
presumption that the ones that remained inherently
must have been good and that the platform needed to
have immediately regarded them as being valid when the platform
had significant experience to know that there
were validity problems with the copyright notices. Not just in terms of the claims
but also in terms of the form. The court criticized them for having blacklisted Rice
Corp notices prior to prior to BMG having hired Race Corp. But it didn’t discuss why Race
Corp had been blacklisted. So even if we want
to go to the idea that filesharing is a certain
type of infringing activity and on the full-service ISP,
that’s a certain type of thing and we want to treat
it differently, we still have the abuse
problem built into it. And then if we try to build a
system that ignores that abuse, and then we extrapolate
it out to more of the, more I think the
point you’re making is like social media tends to be more expressive
type of platforms. It’s the same system
that rules them all. Granted, they were in
the 512(c) environment, but if we don’t have
adequate ways of getting rid of the abuse, and yet we still
have these incredible sanctions falling on users who are
expressors, we’ll take out all of it, including the stuff that we think should actually
have been an easier decision.>>Okay, Mr. Gratz.>>Joseph Gratz: I want to begin by answering Ms.
Smith’s question to Ms. Gellis regarding a
continuum, and my answer is yes. That is, a continuum
is critically necessary in the context of
repeater infringer policy. Especially when distinguish, as
we must, between different kinds of service providers,
different kinds of subscribers and different kinds of activity. I think all of those
things go into the question of whether the implementation of a repeater infringement
policy is reasonable, and indeed, ought to go into each individual service
provider’s individual decision making on whether to
terminate a particular user. And in my experience that
the service providers who, in Mr. Feerst’s words,
care, take all those things into account and want to get
rid of abusers but don’t want to trammel either lawful speech
or the non-infringing speech of people who make mistakes. And so, I think what I come
back to here is the difference between conduit ISPs and edge
services in this context. With respect to conduit
ISPs, the continual, one way the continual was
implemented and maybe one model for voluntary implementation of that continuum is
the Cox red alert system who we haven’t heard a
lot about that was a way of doing a graduate response
where the ultimate sanction, the thing at the end of the
road, was a very serious thing. And I think the seriousness of terminating conduit
ISP service is of critical importance. I mean it’s not that
hard to square with the goals of the DMCA. I think the reason on an earlier
panel we heard Packingham mentioned, the reason Packingham
is important is not necessarily because the First
Amendment imposes limitations on Congress’ ability to act
in this area, though it may. I think the reason that
Packingham is important is because it points out
as a policy matter, these other things that
we consider so important as a society, avoiding,
you know, spousal abuse and child abuse, yield to
the necessity of using, of people using the internet
and people who make mistakes and maybe even repeated
mistakes on those same sites. Those interests, at
least in some cases, where it’s appropriate,
where it’s reasonable, also yields to the
necessity of participating in society using the internet.>>So you mentioned
the difference between conduit internet
service providers. They all have the same
knowledge standards, though. But notice the takedown
doesn’t apply to 512(a), although with Cox and Grande
maybe it does apply to 512(a), and they are expected to
viewees keeping track of some of the notices they’re
receiving, though they don’t have
a formal system for it. So I guess the question I have, and I’m also hopeful Mrs. Ray
can respond to this in terms of what her members are doing. But is since Cox, and I
know Grande is too recent to have affected any sort
of practical response, but since Cox, how are conduit
types reconciling the fact that maybe there isn’t that much
of a difference between them and the other types
of service providers in terms of obligations.>>Joseph Gratz: So I don’t
have any information for you about what is happening
on the ground. What I can, all I can add before
I defer to Mrs. Ray is to say that from the point of view,
it still isn’t, even after Cox, a notice and takedown
type regime. That is, notices that are
received area one input to a reasonable determination
whether someone is a habitual abusive infringer.>>Okay, since you were
called out, I’m going to go to Mrs. Ray and then Mr. Feerst. We’ll let you have
the last word.>>Tamber Ray: Thank you. Just quickly I wanted to,
you know, second the notion of Ms. Gellis and Mr.
Schwartz on their comments about whether there’s a need
to forward the actual notices and what the obligations and responsibilities are
on behalf of the ISP. When they forward
those actual notices, especially if they contain
the settlement language like as in Cox. That left a big question
mark for our members after the court decision. Secondly, you know, to Ms.
Gellis’ point about sort of invalid DMCA notices, in particular the question
that’s come up for us is where they’re sent to multiple
email addresses as opposed to just to the DMCA
email address. Can they completely
disregard any emails, any notices that come to an email address other
than their DMCA one? Assuming, of course, they
have a valid operational DMCA email address. But more to you question and
thank you for the opportunity. The members that we’re
seeing, they are concerned with the Grande opinion, recognizing that it’s only
summary judgment phase. But it really did put a pretty
hefty obligation on the ISPs and does not seem to take
into account the concerns that have been raised
here about the number of valid email addresses,
the number of valid notices. The other question concern
that comes up is what do we do with public Wi-Fi that
we have had some success through technology with
seeing a reduced number of copyright infringement
notices or DMCA notices through instances where peer to peer networking
is no longer used. Where there, the members, the
subscribers, are using more of a streaming service to get their videos,
to get their music. That has helped. But where it hasn’t helped is
in the public Wi-Fi context. That they have educated
their members. A lot of times they find that members have
kind of an open Wi-Fi. And when they educate,
I’m sorry, when they educate their
subscribers as to how to close down their public Wi-Fi so that,
you know, people can’t park on the street and use it. That helps until another
one pops up somewhere else. And with, in the case
of Wi-Fi in hotels, Wi-Fie in restaurants,
they are kind of concerned of what are they going
to do about that. So any clarification in that
respect would be helpful. And then finally, you know, to the extent there’s an
expectation obligation of ISPs to disconnect their
subscribers at some point, we need further clarification
and perhaps coordination with a state government, with the Federal
Communications Commission. State governments have
considered internet service as an essential service. Therefore, there’s going to
be certain very concrete steps that would be needed
prior to disconnecting from the FCC standpoint. There are performance
obligations that these ISPs are
required to meet to comply with FCC requirements. So they’re kind of in a box
here, and any further guidance, any absolute clarification
would be helpful. Thank you.>>Okay, thank you. Mr. Feerst.>>Alex Feerst: Yeah, I wanted
to provide a very partial answer to Mr. Greenberg’s question
before on repeat infringer, which is what examples do we
know of, of strategic behavior, of sort of either batching or serially sending infringement
notices and then using that to sort of strategically
try to get a particular
user removed. If you look through
our Lumen notices, I would say I do see these. It’s infrequent, so I don’t want to overstate it It’s a
rarity, but they do have it. And it may happen more
elsewhere, and it may increase. If you look in the Lumen
notices that we upload related to infringement, one of
the very idiosyncratic but interesting cases on this
is an investigative journalism publication right out of Ecuador that was using Flickr
photographs of politicians posted
by the government of Ecuador’s Flickr account. And the government of Ecuador
then sent some copyright takedowns and TMC takedown
notices, along with a request that that account be banned
based on a repeat infringement. So, as I say, that
is at the margins of what we’re talking about, but I think it really suggests
the potential for mischief that can be done with
strategic behavior around these repeated infringer
policies from my perspective, the risk and the lack of clarity
around the necessity of going to summary judgment to have
it evaluated whether the way that we responded to that was
appropriate is a six-figure cost that is gigantic for a startup. So we very much want to do the
right thing, and it’s very clear that that’s a form of
political speech that we want to be very careful about. But at least as to how
repeat infringer policy works in the existing cases, there
wasn’t really any guidance for me to de-risk that in
a way that was meaningful. Thanks.>>Okay, thank you. So with that, we’ve
come to the end of the domestic portion
of the day. I just want to remind everyone, there is going to
be an open mike. If you have additional comments
or other concerns that you want to raise that we didn’t get to,
please feel free to sign up so that you can speak during that. We’re going to take a 15-minute
break, and then we’ll be back with the international
panel at 2:45.


Leave a Reply

Your email address will not be published. Required fields are marked *