Judges’ Queries and Presenter’s Replies

  • Icon for: Jeffrey Lidz

    Jeffrey Lidz

    Judge
    May 21, 2013 | 03:35 p.m.

    How can this research contribute to stopping internet piracy?

  • Icon for: Zubin Jelveh

    Zubin Jelveh

    Presenter
    May 21, 2013 | 05:54 p.m.

    Hi Jeffrey,
    Thank you for your question, this is a decidedly tough problem faced by the producers of digital goods. Two of the main approaches taken to preventing piracy are legal actions (e.g. RIAA lawsuits or France’s HADOPI legislation) and technological impediments (e.g. DRM). There has been little evidence to date that either of these two approaches has had much success in preventing P2P piracy, but our research suggests that legal entities could play a role in mitigating piracy on cyberlockers.

    For example, our research identifies that the volume of infringing content available on cyberlockers can be manipulated by the presence and design of incentive schemes. Taken to its extreme, that would mean if cyberlockers were prevented from offering payments or rewards to uploaders (or, more realisticly, forced to create a more robust screening system for monitoring infringing content), then we would expect to see less piracy. And we did find this to be the case with RapidShare, a major cyberlocker that stopped its rewards after legal pressure.

    A big hurdle here, however, is that many of the cyberlockers operate in regions where pressure from authorities in the U.S. and Europe have less influence. Still, the shutdown of Hong Kong-based MegaUpload, one of the biggest cyberlockers and among the first to reward uploaders, had a huge impact on the cyberlocker ecosystem. Many of the cyberlockers that vanished were the ones offering incentives, and traffic analysis following the shutdown suggested that the volume of content on cyberlockers fell compared to P2P volume. This points to the fact that, relatively speaking, cyberlockers are more vulnerable to concerted efforts by legal authorities than bitTorrent because the cyberlockers themselves are points of failure, whereas points of failure in P2P systems are harder to identify.

  • May 21, 2013 | 09:41 p.m.

    Hi Zubin,
    Could you discuss the sampling bias you feel is present in the earnings data? Do you feel that people over-report what they earned, under-report, or just a lot of people don’t report publicly?

  • Icon for: Zubin Jelveh

    Zubin Jelveh

    Presenter
    May 22, 2013 | 11:16 p.m.

    Hi Aurora,
    Given how popular these sites have become in the last few years, it’s almost certain that the vast majority don’t report publicly. As for those who do report, the earnings distribution seems to suggest that over-reporting is not rampant. The primary reason for this is that if we expect people to over-report, then there should be fewer people reporting low earnings, but we found that a large majority of posters earned less than $15 per day (and a majority even less than that).

    As for under-reporting — which could happen to give a false impression to law enforcement, or competing uploaders, about the economic opportunities present — we conjecture that based on the text accompanying the screenshots, it did not seem under-reporting would be a large problem either. The uploaders that were not earning very much tended to ask for help from others, and the ones who were earning more than average sometimes boasted about their prowess.

    Having said this, without knowing much about the “ground-truth”, these are only educated guesses regarding reporting behavior.

  • Icon for: Sandra Pinel

    Sandra Pinel

    Judge
    May 21, 2013 | 09:48 p.m.

    How might you take the next step to research the feasibility of different cyber locker incentives and disincentives if the goal is to avoid piracy?

  • Icon for: Zubin Jelveh

    Zubin Jelveh

    Presenter
    May 22, 2013 | 11:36 p.m.

    Hello Sandra,
    This is a tough problem indeed: The downloaders are coming to these sites to acquire copy-righted content (since if it wasn’t copy-righted it could most likely be found for free on a site like YouTube). Therefore, while changing uploader incentives to reduce the likelihood of sharing copy-righted content (by stricter monitoring or increasing rewards for non-infringing file) would reduce piracy, there is little incentive for the cyberlockers themselves to enact these changes since it would harm their bottom lines.
    However, there are a few major cyberlockers that (appear to) want to limit the amount of piracy that takes place as a result of their sites (e.g. Mediafire, and 4shared). One thing that sets these guys apart is that they don’t offer uploader payments or rewards. Also, a number of cyberlockers ended their incentive schemes after legal pressures. This suggests that the incentives which need to be altered are at the cyberlocker level and not the uploader level. One possible research path could be to examine the characteristics of the enforcement threats and/or actions that cause cyberlockers to alter their incentive schemes. (Spillover effects are also important. Do certain types of legal actions cause more than just the target cyberlocker to change their business model?)

  • Icon for: Wayde Morse

    Wayde Morse

    Judge
    May 22, 2013 | 12:17 a.m.

    It might be interesting to address how the legal frameworks have/are adapting to the piracy in cyberlockers.

  • Icon for: Zubin Jelveh

    Zubin Jelveh

    Presenter
    May 22, 2013 | 11:47 p.m.

    Hi Wayde,
    I think this is a keen insight. As I mention in my responses to Jeffrey and Sandra, cyberlockers appear to be particularly vulnerable to the action of legal authorities (compared to P2P file sharing systems).
    Still — using the MegaUpload shutdown as an example — where these cyberlockers are only legally vulnerable if it’s shown that owners had knowledge of the infringing activity that was taking place on their site, and the encouraging of such activity. Thus, it’s theoretically possible that a cyberlocker would not be found liable if they could show in court that there is doubt that the owners had knowledge of such activity. This is similar in spirit to Google not getting into trouble for linking to torrent/piracy sites. Unfortunately, at this point the recent legal responses in the U.S. (e.g. SOPA) have been viewed as being overly heavy-handed. And the protection that is already in place (such as DCMA) puts all the onus on the copyright holder.

  • Icon for: Gary Kofinas

    Gary Kofinas

    Judge
    May 22, 2013 | 12:35 a.m.

    Zubin – Fascinating work! I’m curious to hear more about the incentives of the people who host these sites? Are they earning a cut in the flow of cash making it worth their while in significant ways?

  • Icon for: Zubin Jelveh

    Zubin Jelveh

    Presenter
    May 23, 2013 | 12:00 a.m.

    Thanks Gary!
    There IS evidence that major cyberlockers generated millions in revenue, primarily from infringing content. In fact, it’s likely that the cyberlockers are earning much more than the uploaders. The “genius” of the cyberlocker business model is that they figured out a way to let uploaders participate in (a likely quite small) part of the companies’ revenues. One eye-popping datapoint that came out of the U.S. shutdown of MegaUpload was that the company paid $35 million to a hosting service in the U.S. It’s quite likely that this was financed by premium accounts that downloaders purchased from the cyberlockers, and also advertising revenue. It’s worth noting that when a user downloads a file from these sites, there is typically an ad shown on the download page. The exchanges which serve these ads are run by major companies whose names we are all familiar with. In their heyday, there were a number of cyberlockers which were among the 100 most visited sites in the world. This suggests that these ad exchanges also earned revenue from the cyberlockers.

  • Icon for: Gary Kofinas

    Gary Kofinas

    Judge
    May 23, 2013 | 07:21 p.m.

    Thanks for the thoughtful response. I;m impressed. g

  • Further posting is closed as the competition has ended.

Presentation Discussion

  • Icon for: Zak Gezon

    Zak Gezon

    Trainee
    May 23, 2013 | 09:24 a.m.

    This is interesting, and also troubling. Given the nature of the internet, I am curious what the solution is, or if you feel that there needs to be a solution?

  • Icon for: Zubin Jelveh

    Zubin Jelveh

    Presenter
    May 24, 2013 | 11:25 a.m.

    Hi Zak, Thank you for the comment. I think there needs to be some kind of solution, but there a number of different angles to consider.
    For example, something doesn’t feel right about companies like cyberlockers (or ad exchanges or search engines) making money off of copy-righted content when the authors themselves can’t. (A feature of cyberlockers, but much less so in P2P.)

    Also, piracy does seem to serve the purpose of sampling. You don’t know if you’d like a certain movie, book, or song. You get an illegal copy and after a couple of minutes watching/listening/reading/, it turns out you don’t like it. No harm, no foul (right?).

    So solutions need to balance both the rights of producers and the habits and expectations of consumers. While legal approaches have shown some impact on reducing piracy (e.g. HADOPI in France), my sense is that business model innovations will have a greater impact. (e.g. the freemium model for mobile apps).

  • Icon for: Ashley Richter

    Ashley Richter

    Trainee
    May 23, 2013 | 03:45 p.m.

    Great tagline on the video! Pertinent research in today’s digital society- great work! One wonders what your thoughts are with reference to the ethics of the computer piracy and how they are/will/won’t shape legislative policies and cultural climates of the future?

    Really great work guys!

  • Icon for: Zubin Jelveh

    Zubin Jelveh

    Presenter
    May 24, 2013 | 11:40 a.m.

    Hi Ashley, I’d love to take credit for the tagline, but I found it online. I think one of the main problems with legal approaches is that they are relatively simple to circumvent and promote an ‘arms-race’ in detection and evasion. For example, there is a new “six strikes” policy which was an agreement between the ISPs and the movie industry. The way it works is that ISPs will now monitor their traffic for downloading of infringing content, and after each violation a progressively harsher penalty will be imposed on the ISP subscriber. However, there’s been a boost in demand for VPN services and proxies that allow ISP subscribers to hide their traffic and circumvent the monitoring. The worst possible outcome of this could be that the money spent in detection and evasion is greater than the money lost to producers from illegal file-sharing.
    As for the influence of culture/ethics, I don’t think our society has a strong feeling one way or the other about how wrong piracy is (if anything, I’d venture to guess that most people don’t think it ranks high on the list of offensives.) For that reason, and as I mentioned in my reply to Zak, I think the best defense are services that make sampling and paying for these digital experience goods a much easier experience. (e.g. Netflix and Spotify)

  • Icon for: Paul Tanger

    Paul Tanger

    Trainee
    May 23, 2013 | 06:47 p.m.

    wow, really fascinating!

  • Icon for: Zubin Jelveh

    Zubin Jelveh

    Presenter
    May 24, 2013 | 11:08 a.m.

    Thanks paul!

  • Further posting is closed as the competition has ended.

Icon for: Zubin Jelveh

ZUBIN JELVEH

Polytechnic Institute of New York University
Years in Grad School: 2

Profiting from Piracy: A Measurement Study of the Cyberlocker Ecosystem

Piracy has been an ever-present danger for producers of digital goods since the rise of the internet. This ongoing work looks at the pirating ecosystem through a variety of lenses.

For the NSF IGERT video and poster competition, we present results of a measurement study of the economics of cyberlockers. These sites are a new breed of online businesses which store users’ files on remote servers and allow users to access files from any location. The number of cyberlockers has surged since 2006, likely as a function of lowered bandwidth costs and the ease with which cyberlocker services allow users to share copyrighted material.

A central, and troubling, innovation of cyberlockers is that they allow sharers of infringing content to earn money from this activity, in stark contrast to person-to-person platforms like BitTorrent. In order for cyberlockers to become popular, they must entice uploaders to share their links. We examine the payment mechanism by which cyberlockers have been able to incentivize these uploaders. Full Paper