April 24, 2014

avatar

Is SafeMedia a Parody?

[UPDATE (Dec. 2011): I wrote the post below a few years ago. SafeMedia's website and product offerings have changed since then. Please don't interpret this post as a commentary on SafeMedia's current products.]

Peter Eckersley at EFF wrote recently about a new network-filtering company called SafeMedia that claims it can block all copyrighted material in a network. We’ve seen companies like this before and they tend to have the warning signs of security snake oil.

But SafeMedia was new so I decided to look at their website. My reaction was: what a brilliant parody!

The biggest clue is that the company’s detection product is called Clouseau – named for a detective who is not only spectacularly incompetent but also fictional.

The next clue is the outlandish technical claims. Here’s an example:

Pirates are smart and innovative, and so is Clouseau. Our technology is dynamic, sees through all multi-layered encryptions, adaptively analyzes network patterns and constantly updates itself. Packet examinations are noninvasive and infallible. There are no false positives.

Sees through all encryption? Even our best intelligence agencies don’t make that claim. Perhaps that’s because the intelligence agencies know about provably unbreakable encryption.

Wait a minute, you may be saying. Perhaps SafeMedia was just making the usual exaggeration, implying that they can stop all bad traffic when what they really mean is that they can stop the most common, obvious kinds of bad traffic. Good guess – that’s the usual fallback position for companies like this – but SafeMedia doesn’t shrink from the most outlandish claims of infallibility:

What if illegal P2P no longer worked? What if, no matter how intelligent, devious, or well-funded an Internet pirate was, they absolutely could not transmit copyrighted material via P2P? SafeMedia’s goal was to create the technology that would achieve exactly this. And we succeeded.

Employing our new technology, Clouseau and Windows + Transport Control, makes illegal P2P transmission of copyrighted material impossible. IMPOSSIBLE. Not difficult and not improbable. IMPOSSIBLE!

The next clue that SafeMedia is a parody is the site’s blatant rent-seeking. There’s even a special page for lawmakers that starts with over-the-top rhetoric about P2P (“America is at war here at home within our own borders. And we are taking casualties. Women, men, and children.”) and ends by asking the U.S. government to act as SafeMedia’s marketing department:

We need the Congress to pass legislation appropriating funds for installing the technology on every Federally-supported computer network in the country, most importantly in educational institutions (schools, colleges, universities, libraries)…. We need the Department of Commerce to promote using the technology in all American businesses big and small, and to push for its international adoption. We need the Department of Education to insure that every educational institution in the USA, private and public, primary and secondary, college and university, is obeying the law.

You now have the right weapons. Let’s end the war!

Add up all this, plus the overdesigned home page that makes maddening fingers-on-a-blackboard noises when you mouse over its main menu area, and the verdict is clear: this is a parody.

Yet SafeMedia appears to be real. The CEO appears to be a real guy who has done a few e-commerce startups. The site has more detailed help-wanted ads than any parodist would bother with. According to the Internet Archive, the site has been around for a while. And most convincingly of all, an expensive DC law firm has registered as a lobbyist for SafeMedia.

So SafeMedia really exists and company management thought it a good idea to set up a parody-simulating website and name their product Clouseau. What an entertaining world we live in.

(Thanks to Peter Eckersley for sharing the results of his un-Clouseau-ish investigation of SafeMedia’s existence.)

Comments

  1. Brian (a different one, honest) says:

    So…what is the possible motivation for posting such a parody? I’m just not seeing it. It doesn’t seem like a particularly effective way to argue to the music/movie industry that blocking illegal transfers is impractically hard (in the same way that finding and proving an optimal solution to a 50-city traveling salesman problem is “impractically” hard). Mostly it just seems like something that people like us can go to and get a chuckle out of – and for that it seems needlessly complicated and expensive.

  2. Chris S says:

    Neat. Look deeper, and I suspect that I can paraphrase what they are saying:

    “Let us be the gatekeeper on every PC and on every sub-net, identify to us all your information, and we can track it where ever it goes, and stop it from going where it shouldn’t.”

    This is absolutely *possible*, but politically, technically, financially, and socially infeasible.

    Yes – they could do it, but -
    – it would require laws that nobody would pass
    – it would require technical acceptance that many will not give
    – it would require expenditures we simply can’t afford
    – it would require us accepting an intrusiveness we can’t accept

    Given how hard is has been to date to explain to many governments and content industries just how infeasible their dreams are, I think this is a nice twist on educational outreach.

    Think of it as “erudio ad absurdum”.

  3. Mark says:

    I wonder how they deal with RFC 1149 traffic.

  4. EJTower says:

    Perhaps this is something like a long term activist prank like what The Yes Men do? They have done many big activist pranks like show up on CNN posing as the executives of certain chemical companies, and present to the WTO.

    Just a thought.

  5. Rob says:

    Does anyone think that they could find an image of the router used to make SafeMedia’s (only, that I can find) image of their product? Most Cisco/Linksys product images I could find are similar, but have a slightly different pattern of ventilation holes down the right hand side. My guess is that somewhere on the internet is precisely the same product image, but with Linksys or something in the position of the Clouseau image.

  6. Robin Blume-Kohout says:

    Ed, your post leaves me still confused (this rarely happens!). You’re saying that: (a) SafeMedia is a real company, that really does network filtering; but (b) their website appears to be a brilliant, vicious parody. Should we infer that:
    (1) they’re smart, creative people, and their weird but potentially effective marketing strategy involves parody? or
    (2) they’re woefully misinformed about technology, and really believe this stuff? or
    (3) while their engineers might be competent, their marketing department is out in La-La Land.

    The recent case of D-Wave Systems lends some credence to option (3). D-Wave claims to have built “the world’s first commercially viable quantum computer.” The academic quantum computing community (full disclosure: that includes me) responded with near-universal skepticism.

    Some of the reasons are technical, but one isn’t. They’ve marketed their device as an efficient solver of NP-complete problems. This is widely believed (with fairly strong evidence) to be impossible — we think quantum computers can’t efficiently solve NP-complete problems.

    There are some pretty good people working there, and they’ve got a good shot at building something that improves on current approximation heuristics for Max-Clique (and possibly some other problems). Unfortunately, they’ve alienated much of the academic community with deceptive marketing claims. If they crash, we don’t want to be tarred with the same brush.

    Sadly, I get the impression that aggressive marketing is par for the course in new-tech startups. Could this be the case with SafeMedia, or is it just too over-the-top?

    Oh… and, Brian, I think you chose a bad example to make your [valid!] point. AFAIK, TSP is actually tractable up to at least 1000 cities. I’m not sure which NP-complete problem would be better (matrix permanent? max-clique?).

  7. QrazyQat says:

    To paraphrase AC Clarke, any sufficiently advanced marketing is indistinguishable from parody.

  8. Bob Monsour says:

    My favorite statement about the product is:

    “Many advanced technologies have been deployed to achieve that goal, such as symmetric multi-processing (up to 16 cores), hyper-threading, network polling, and Deterministic Finite Automa (DFA).”

    Impressive indeed ;-)

    Regards,
    -Bob

  9. dmc says:

    The funniest thing is that (at least parts of) the recording industry seem to be taking it seriously, not recognizing either outlandish parody or outlandish snake oil. Googling for clouseau and safemedia brings up a lot of reports.

  10. Neo says:

    What I’ve read about quantum computing indicates that it could efficiently solve NP problems, by taking every possible answer, running the polynomial-time verification on them all in parallel (using superposition), and returning the answer(s) that verify (as a superposition, if more than one). That just means rigging the apparatus so that verification failure results in self-cancellation. E.g. the verification algorithm actually inverts the phase of its input for “no”, but not for “yes”, and the input and output are summed — the wave function in can be an equal superposition of all the inputs, and the wave function out will contain non-zero components only for the ones that “pass”. That solves your NP problem in polynomial time. Ones with no polynomial-time verification obviously don’t work, but e.g. factoring of large numbers becomes efficient. If you have enough qubits to hold the answer plus whatever temporaries the verification algorithm needs, of course — for a problem with 2^n possible answers, you’d need a minimum of n qubits.

    Has something emerged in the research that makes this difficult, yet doesn’t torpedo the whole idea of quantum computing? (Inability to preserve correlations among the qubits would seem to render the whole idea pointless, for instance.)

  11. Conrad says:

    This looks like a brilliant plan to take money from the media elite who are too stupid to recognize the site as parody. Have them sign a 1-year contract, get them to pay up front, then snip their internet wires… That’s the only way to truly guarantee that no copyrighted content ever gets through!

  12. Ali says:

    They didn’t say anything about false negatives. Maybe it’s the internet version of the paper clip except it uses a bearable character from an awesome movie.

    My first thought was what Conrad said, but I guess it depends on what you mean by positive.

  13. Silas says:

    @Neo
    Actually, the definition of NP-Completeness requires the problem to be in NP, which is defined as the set of all problems that, if you guessed a solution, you could verify it in polynomial time.

  14. PO8 says:

    Neo Says: “What I’ve read about quantum computing indicates that it could efficiently solve NP problems, by taking every possible answer, running the polynomial-time verification on them all in parallel (using superposition), and returning the answer(s) that verify (as a superposition, if more than one).”

    I’ve read that a lot too. However (assuming that what you mean by “NP problems” is “NP-complete problems” or “NP-hard problems”) it probably isn’t true. As far as we know, not just any problem can be formulated in such a way that it can be verified using quantum superposition. The current belief among theorists is that QP

  15. PO8 says:

    (sorry—html) that QP < NP. In other words, quantum computers can’t solve NP complete problems in polytime.

  16. Adam Ierymenko says:

    It’s a scam, and a brilliant one. It’s basically a dowsing rod. Go to http://www.randi.org and search for dowsing rods– Randi has an extensive archive of outlandish claims for dowsing rods to detect everything from oil to drugs in students’ lockers with all kinds of pseudo-technical “quantum field effect” hooey in their marketing.

    This is basically a dowsing rod for packets. It will probably employ some basic P2P network detection and blocking heuristics and then make all sorts of really outlandish dowsing-rod-style claims. The goal is to part clueless buggy whip makers (the music industry) with their money, possibly using the government as well.

  17. Lawrence D'Oliveiro says:

    At last! The long-sought-after implementation of RFC3751!

  18. Researcher says:

    On their Company Contacts, and About Us page, they list a number of phone numbers. Some of phone number’s match up, but other’s are suspious.

    http://www.google.com/search?q=561)-750-5099&hl=en&pb=r&sa=X&oi=rwp&ct=title
    gives an address for “Joseph L Maher III” of 2880 NW Boca Raton Blvd,Boca Raton, FL 33431, which is the name given for that phone number.

    (561) 989-1934 is probably their office phone, because that is the number for Sidney Klein, Vice President, Strategic Alliances; Summer R. Davis, Communications Specialist, and (according to their whois record), Marcos Zepeda, listed as their Administrative Contact, with an address of 2500 Ave Au Soleil / Gulf Straem, FL 33483 (Yes, it is spelled Straem in the whois record)

    The office address seems to be: 6531 Park Of Commerce Suite 180 / Boca Raton, FL 33487

    What’s most odd, I find, about the webisite, is the lack of any clear way to actually *buy* anything off it. There are no prices, there is no where even specifically set up to put your email address in and have them send you prices. I guess they may be not aiming at “individuals” but it seems rather odd.

  19. kaukomieli says:

    hm, maybe they have developed some kind of windows-packet-driver that has to be installed on every government-owned machines. how they intend to identify the packet-load of traffic that gets encrypted by an application before it goes on the wire is beyond me.

    maybe bittorrent-packets (and the like) have identifiable patterns even when encrypted?

    If one wants to stop all p2p-traffic by installing a software on the users computer (who has to be a restricted user and not local administrator…) this could be possible, since it would not be necessary to know the content. This would kill all legitimate p2p-use, but on government-computers i can’t see that anyway.

  20. FBOS Detective says:

    Is SafeMedia, a Parody?…not literally, as I am sure the author intended to ask rhetorically since the company appears to be real, but publicly and proactively maintains claims that are simply out of this world. SafeMedia appears to be a marriage between Dr. Strangelove and Professor Harold Hill. A sociopathic technologist combined with slick-tongued over-the-top salesman.

    SafeMedia is obviously in the business of selling FBOS (Flaming Bag of S#!T) technology as evidenced by:

    - Absolute Claims: Extreme hyperbolic marketing is par for the course for technology start-ups, but claims of omniscience and infallibility are the calling cards of outright fools.

    - Crazy Business Model: PR & Lobbying to drive demand, with zero explicit support from anyone or any organization from rights holders to legislators.

    - No Proof of Technology:, Any available technical specifications, whitepapers, or independent benchmarking? No…and not bloody likely any time soon.

    *No matter your opinion of rights holder trade association, the technical people within possess more than enough savvy and influence to call BS on bad technology. Furthermore, they are not in a position to buy, just vet; and benchmark processes are in place before any recommendations are made. Nonetheless, the point at which those recommendations feel like mandates is what many people get uncomfortable with, especially when they are infused in the lobbying efforts of the rights holder industries. However, SafeMedia seems to putting all of its effort to assumptively bypass all of that, with little more than lip service to proof of concept: Clouseau is now being successfully used in Florida, California, Oklahoma and Texas…”

    Maybe not a parody itself, but perhaps a parody on the disclaimer seen in the end of many movie credits: “”The events depicted by company this are fictitious. Any similarity to reality, living or dead is merely coincidental.” would best describe SafeMedia.

    Anybody feel like singing “Shipoopi”?

  21. FBOS Detective says:

    Got to love this press release:

    http://www.pr.com/press-release/36553

    “SafeMedia Corp. Technology Signs MAYO Communications for National Publicity and Education Campaign

    “Sales of pirated music CDs were worth an estimated $4.5 billion and there were about 20 billion illegal downloads,” said one study. SafeMedia Corporation, a Boca Raton, Florida based technology developer and 80-Year Old Comedian retain MAYO Communications for PR.”

    Nice two-fer!

    If you want another good chuckle check out the ultra modern (circa 1995) website from MAYO Communications:

    http://www.mayocommunications.com/

    the begining may fool you..but stick with it & you’ll see what I mean

    -FD

  22. Robin Blume-Kohout says:

    Neo, PO8,

    Sorry for not responding earlier — forgot to lurk!

    Regarding why quantum computers can’t solve NP-complete problems, the best resource right now is Scott Aaronson’s blog. I suggest as a starting point. The proof that you _can’t_ solve NP-complete problems efficiently via “quantum parallelism” is called the BBBV theorem, and that article has a link. I also recommend , which explains how Shor’s algorithm _does_ work to factor integers. It makes heavy use of the structure in the problem.

    Finally, I have to point out that BQP (the class of problems a QC can solve efficiently) is _not_ thought to be within NP. The two classes overlap, but each contains problems outside the other. For instance, simulating quantum systems is what a quantum computer does best — but it’s not in NP, because a classical computer can’t even check the answer!

  23. Neo says:

    “…and that article has a link.”

    Unfortunately, though, your blog post doesn’t, and phrases like “Scott Aaronson’s blog” are not meaningful to today’s DNS resolution servers. And making them so would appear to be AI-complete…

  24. Neo says:

    Oh, and “I also recommend , which explains how Shor’s algorithm _does_ work to factor integers.” certainly isn’t very helpful. You also recommend what, exactly? Apparently its name is a single space character, which is not going to make it at all easy to track down…whatever it is.

  25. Robin Blume-Kohout says:

    Neo,

    I’m awfully sorry about that — but WordPress decided, for some reason that I can’t intuit, to strip out my URLs. Perhaps I used brackets incorrectly (and got them interpreted as HTML), or perhaps I’m simply not authorized to post links.

    Anyway, here goes again. Freedom-to-tinker doesn’t have a preview function for comments, so I’m shooting in the dark.

    Scott’s blog, “Shtetl-Optimized,” can be found at http://scottaaronson.com/blog . The post that explains why we don’t think “quantum parallelism” provides a speedup for NP-complete problems is http://scottaaronson.com/blog/?p=206 , and the explication of Shor’s algorithm is http://scottaaronson.com/blog/?p=208 .

    My apologies to Ed for cluttering up his blog, and to everyone else for the link failure! If this fails for some reason, I’m giving up.

  26. Kaz says:

    Robin,

    bbbv simply indicates that govers algorithm is optimal. Although a O(log n) search implies P=NP wrt quantum computers. The converse does not hold. That is, bbbv does not provide a proof that “quantum parallelism” cannot be used to solve NP complete. bbbv ignores the structure of NP complete problems and simply focus on the search problem which can be used as a “brute force” method to solving NP complete problems in super-polynomial time.

    Cheers

  27. Anonymous says:

    I was wondering if

    http://www.aacsla.com/home

    is a parody site?

  28. c0uchw4rrior says:

    Apparently SafeMedia’s CEO recently testified in from of the US Congress regarding piracy and its technolgies. The event was featured on Slashdot.org today, http://politics.slashdot.org/article.pl?sid=07/06/06/1652254

    Can anyone say “perjury” ?

  29. Nobody says:

    I just figured out how it works! It blocks *all* bad content, right? I bet they just cut your network cable and call it a day!

    Uhoh. I hope I haven’t accidentally guessed someone’s trade secret :-)

  30. Kamil says:

    A disturbing twist to the tale:

    http://politics.slashdot.org/article.pl?sid=07/06/06/1652254

    Apparently they weren’t joking. I mean, no one lies to Congress, right?

    PS: Way to rip up the dancefloor at reunions!

  31. Epic says:

    I used to work for Safemedia, and rest assured that they are real and the product is functioning fine. Time will tell.

    It is currently installed in a few universities and performing as they have stated to the public.

  32. Ed Felten says:

    Epic,

    Which universities are using it?

  33. Doorhand says:

    The software disables p2p networks. Read up! If they can not connect to other peers there will be no sharing. Although your right marketing does a crap job.

  34. XRAY says:

    It is a LIE. We installed it in our campus and every 10 minutes we had a complain. We decided to return the unit. They don’t have any technical support either. It is a horrible thing to have ….on network.

  35. JackRing says:

    What exactly does this product accomplish? If the product simply disables protocols at layer 2, then any legitimate services that were carried with those protocols go away as well. How is it different from a network gateway with analed off services?
    None of the advertising contains any details about what the product is actually doing. Does it do any inspection of the media which is being exchanged? Or just the protocol header?

  36. Bill Weiss says:

    Infringing P2P client networks have very specific characteristics, such as inadvertent file sharing, immediate uploads for downloaded content, an overlay routing layer and no input / output control. Excluding bit torrent from this discussion, a good technology (like SafeMedia) can definitely separate and segregate infringing P2P networks.

    As you know from our website, SafeMedia encourages the use of bit torrent and established the safe torrent site to help users obtain non–infringing torrents which will not include any infringing material. As a technologist, you can discover the content of any torrent thru the seeds and trackers associated with the torrents.

    SafeMedia doesn’t stop all P2P networks. Many p2p networks, such as ruckus, lion share, voip and streaming pass freely. The objective of SafeMedia’s system is to protect against identity theft, reclaim network bandwidth and reduce digital infringement.

  37. Bill Weiss says:

    In response to what XRAY wrote. The complains you received were from students who couldn’t continue to steal content. We have no units returned from any school or business.

    BLOG truthfully!

  38. JackRing says:

    You contend that the way to stop the malicious exchange of copyrighted materials is to disable the common protocols that mediate them? I could care less what some third rate south of the mason dixon company “encourages or discourages.” There are many MANY uses of P2P networks besides “stealing content.”

  39. Bill Weiss says:

    We agree that there are MANY uses of P2P besides content piracy. I outlined our open access for many P2P clients including hosting of open source on Safe Torrents.

    Your understanding about common protocols lacks credence. Please look closely at inadvertent file sharing which is adopted widely by infringing P2P.

    Honesty from south of the border is still better than dishonesty from north of the border

  40. X-RAY says:

    Of course … Bill
    You won’t consider a Pilot to Deployment denial as refusal or returned unit.
    But open your eyes to the fact that since 2006 you have sold less than 10
    units based on your false claims. Any way sooner or later time will open your
    eyes….. And by the way how do you keep up to date with changing P2P
    clients ???? Market Truthfully……

  41. Xa-Sun says:

    Looks like this company is already dead.
    If it is active please someone blog because we are launching a P2P software project and we will definitely put them to death. what is the pleasure in paradise if you don’t have freedom??
    Join with us to eradicate SafeMedia — and all of its kinds

  42. Anonymous says:

    The founder is the greatest lier who hires different people saying he will give bonuses when he doesn’t have money to pay full paychecks. Why can’t people do enough research before joining this company ?

  43. Tom says:

    For X-Ray: Posting; For Xa-Sun and For the Anonymous posting: You are fighting a losing war. Attacking the company or the founder will not solve your problem, for your information freedom does not mean infringing on other people legal rights.
    I suggest that all of you read: http://arstechnica.com/news.ars/post/20080801-college-funding-bill-passed-with-anti-p2p-provisions-intact.html “College funding bill passed with anti-P2P provisions intact” maybe this will educate you on the law of the land…

  44. Anonymous says:

    By Passing bills you are not going to get business, Buddy.
    Lets see how many days you survive all fragile marketing claims.
    Let the war begin… you will see what is coming ……

  45. X-Ray says:

    Looks like someone cut off Safemedia’s balls. :-D

  46. Anonymous says:

    Pop goes the weesel!

  47. Anonymous says:

    the only thing which is not safe in this world is Safemedia’s future.
    with all these new p2p clients — they still claim to be only 600.
    Who will update it with 0 technical staff and support ????
    out of their F***ing mind. Royal Road to Failure…..