Columns

Child Pornography Blocking Plan a Risk Worth Taking

My weekly Law Bytes column (Toronto Star version, homepage version) wades back into the Project Cleanfeed Canada debate.  My last post on this issue generated considerable discussion with many valid criticisms of the ISP plans to block access to child pornography.  In developing this column, I posed many of the criticisms to Cybertip.ca and found that there were good answers to many concerns along with a willingness to address other issues.  In particular, Cybertip.ca: 

  • is amenable to incorporating judicial review of the block list
  • will issue regular public reports that identify the number of sites on the block list along with the number of participating ISPs
  • will only block images of pre-pubescent child pornography, thereby limiting the prospect of overblocking legal content
  • is developing an appellate process that includes review from the National Child Exploitation Coordination Centre
  • is subject to several internal accountability mechanisms including an independent board of directors and pressure from funders (60 percent of its funding comes from the federal government) to avoid potential liability by overblocking

The column expands on these and other issues.  I should also note that Cybertip.ca was open to further changes in response to suggestions from the Internet community.

18 Comments

  1. Fill in the blank
    ‘Fill in name’, the initiative’s Executive Director, sought to address many of those concerns in an interview last week. She/He noted that ‘initiative I am pedaling ‘, has instituted several measures to reduce the likelihood that the list/law/regulation will extend beyond clear cases of ‘moral repugnant/criminal behaviour/acts’.

    Please note that the paragraphy reads well regardless of wether we are talking about child porn, same sex marriage, gun use, drug abuse, spousal abuse, discrimination, etc.

  2. Can’t cybertips.ca simply come out and say ‘child porn is such a heinous act, that we will throw open our doors to any legimate request to prove that we are totally clean and above board’ is a price we are willing to pay to achieve our goals?

  3. Paul C. Bryan says:

    Still too many unanswered questions
    Michael:

    It sounds like you’ve done some homework, and are more comfortable with the limits that Cybertip.ca is putting in place to provide accountability and transparency.

    You have allayed *some* of my concerns. I’m all for stopping child pornography as long as it does not unduly restrict the rights of law-abiding citizens.

    Where can I read Cybertip’s policy statement?

    Perhaps I can take comfort that you yourself have investigated and are satisfied, but there are still unanswered questions, which in my opinion must be answered before we should contemplate trusting an organization with restricting the freedom of our communication:

    1. Collateral Damage

    In your latest article, you mention that only “sites with images of pre-pubescent pornography” will be blocked. The word “site” is a very broad term.

    If this is IP-based blocking (and I presume it is), then “site” means IP address. Often, many sites share the same IP address. Many ISPs provide “virtual web sites”, multiple web sites, each maintained by a different author, sharing the same IP address.

    Presumably, once blocked, all sites on the same IP address go off the air, regardless of the nature of any of the other sites’ content. They become the collateral damage, I presume?

    I’m willing to bet that Google (at least transiently) caches offending content from time to time. Nobody’s going to block Google, right? So, some kind of discretion is applied, right? How is discretion applied?

    2. Appeal

    There is still the nagging question that if the list is secret, how does an (especially innocent) site author get to know his or her site is being blocked? Presumably, one must know one is listed before being able to appeal one’s being listed in the first place.

    Once an author (somehow) determines their site is blocked because of Cleanfeed, how do they apply to have their site unblocked?

    Does Cybertip.ca provide any deference to those who are collateral damage in their battle on child pornography?

    What further recourse does a site author have if an appeal is denied (unilaterally) by Cybertip.ca?

    How often are blocks reviewed? This would be especially important to limit collateral damage. Presumably, once a child pornographer’s site has been blocked, the pornographer will change addresses. This results in the continued blockage of legitimate sites.

    3. “The Data”

    What is the statistical basis of the estimates of damage that Cleanfeed is intended to protect against (“numerous anti-child pornography organizations confirm that there remain hundreds of child pornography sites that may be inadvertently accessed thousands of times each day by unsuspecting Internet users”)? You’ve presumably seen the data and weighed the balance of the saluatory and deleterious effects of Cleanfeed. I’d like to see the data too.

    You stated: “Indeed, earlier this year British Telecom reported that it was blocking 35,000 child pornography hits per day.” How many of these hits were to sites that shared the same IP address, and did not provide offensive content?

    Conclusion

    There are still many questions to be answered. I am not so quick to embrace Cleanfeed without some assurances that innocent people will not be impacted and the process of accountability and transparency have been established.

    The next step in this dialogue would be for Cybertip.ca to get on the record. I see no published policy statement, no open dialogue with the community (unless we’re counting their advocacy by proxy — you).

    Bottom line: we need facts (statistical data corroborating their claims, not just their own estimates) and published policy. Until such information is furnished, I default to maintain status quo, and interpret this as (well intentioned, but ill executed) attack on freedom of expression.

  4. Darryl Moore says:

    Still not convinced
    Please excuse me Michael if I am still a little skeptical and worried about the protection of basic freedoms. Here are some questions regarding your points in order.

    1) Will the judicial oversight be executed before or after a site is added to the list. If it is after then we’ve created a system of presumed guilt and it falls upon the accused to prove his freedom. Sort of like ‘notice and takedown’ no?

    2) What use is the number of sites blocked if we have no way to qualify them? What sort of context can we make use of these figures?

    3) “Will only block images of pre-pubecent child pornography.”, well that’s a motherhood statement. Unfortunately authorities can have a very wide interpretation of this statement. [ link ] . This really does nothing to prevent abuse.

    4) This is probably good.

    5) Your statement presumes that the Federal Government has a vested interest in insuring there is no overblocking. Unless the overblocking becomes extreme and absurd to the point that a significant proportion of the voting public is affected, this will not be the case. The government will be a disinterested 3rd party. I think if you want the government to take an active interest in it then you should make this whole project the responsibility of the government.

    I am still not convinced that this is a good project in the first place. Mostly because I am not convinced that there is a very big problem with this material on the Internet in the first place. With the possible consequences for freedom of speech and the slippery slope arguments about what will be next, I think the the cure might be worse then the disease.

    If a project like this is to go ahead, I am still firmly of the opinion that it must be done through a public and accountable institution, not private corporations.

  5. Matthew Skala says:

    My interpretation of “the benefit of the doubt” is that it’s about burden of proof, meaning something like “we can’t be absolutely sure whether this is a good thing or a bad thing, so we’ll assume it’s acceptable until we have positive proof to the contrary.” For instance, someone accused of a crime might be acquitted on the benefit of the doubt, because there wasn’t sufficient evidence to prove guilt and it’s everyone’s right to have the default assumption be one of innocence.

    For limitations on speech or access to it, I think the doubt goes the other way. Such limitations are almost invariably very bad; that is the default assumption which needs very solid evidence to be rejected; and such limitations are so dangerous that we dare not run the risk of accepting a bad one. So limitations on speech or access to it should be assumed to be *unacceptable* until positively proven otherwise. They aren’t people, with rights, and shouldn’t get the benefit of the doubt.

  6. Michael, it sounds like you’re having a real affect on this issue.

    The incorporation of the judiciary to this process is critical, and with it, this project could do a lot of good.

    Are you planning to work with the government and cleanfeed to develop an open and transparent document that allows a site to be tried before the judiciary with the site information being provided to the judiciary, but sealed from the public, much in the same way publication bans already work?

    It is a fundamental peice of our justice system that says trials must public and that not making them public must have a very very good reason not to. Will every entry on the cleanfeed list have a Cleanfeed VS John Doe Site case available for public review? I would think that there is little harm in releasing a document that talks ABOUT the evidence without actually providing the offending material, that is certified by the judiciary as having been accepted and meeting the rules of evidence.

    How will our rights to be informed of a violation be protected; will cleanfeed give a reference ID to said document upon blocking?

    A 404 — this site is gone, you don’t know why and can’t figgure out if it’s legitimate, seems wholly incorrect. I’m all for this if a nice little -this site has been blocked by cleanfeed under judical order 123523- came up for any blocked content; but otherwise, it still seems to infringe some fundamental principles of justice. It also violates the RFC’s and a 403 — Access Denied with reason as defined in the RFC would make a lot more sense.

    Secret blocking, with secret trials, with secret oversight doesn’t sound Canadian and should correctly be opposed; however, if cleanfeed wants to do this open and notoriously; more power to them. I think the issue comes down to public trust, and we have no ability to trust this organization, or any one branch of government. Our trust is in the system as a whole, and if we’re going to allow this type of thing, then stakeholders from all branches should have to approve. I guarantee if this program was publically reviewed, you’d see a lot of amicus briefs and policy documents that ensure that this program is consistent with the structure of our democracy. Right now it feels like its being created in an end-run around the checks and balances that are so critical to the system.

    There are also common-carrier issues wherein can a site be blocked without notification. If someone has uploaded illegitimate information to a site, and that site is not in the business of moderation, can that site or isp even be blocked without a notice-and-takedown type request. They clearly can’t decide if an image is illegal or not; thats [correctly] not their job as a common carrier. It would seem to me that the presence of material on a network does not necissarily mean that the ISP should be punished and blocked. For example most USENET mirrors could be accused of hosting this type of material, but it is unfeasible to mandate that ISPs monitor all USENET traffic or face an entire blocking of their USENET service. [Don’t give me counter-arguments about they only block websites-blah-blah, give it 12 months). There needs to be some form of notice-and-notice and notice-and-takedown given judicial order that an ISP can comply with in order to ensure that it is not blocked. What about collateral dammage from ip-based blocks? Voices-for-change and the 700 or so blocked sites involved does tend to come to mind….

    In my opinion, cleanfeed should be limited to blocking sites that are a) only outside of Canadian jurrisdiction and where 1) the authorities in the foreign nation have refused to co-operate or 2) have not co-operated within a reasonable time frame (probably something like 60 days) and b) only then, if there is a judicial order saying that there is sufficient evidence to warrant such a blocking.

    Blocking servers within Canadian jurrisdiction is just a bad idea, and it subverts the rights and responsibilities of our justice system. Blocking of Canadian content is NOT within Cleanfeeds jurrisdiction in any way shape or form. This is up to the RCMP to locate and shutdown and we already have the proper tools in place to allow for this type of law enforcement.

    If this were the case, I’m sure that cleanfeed could be both effective and properly overseen. It also bothers me that there are no censorship watchdog groups integrated with this project and willing to give it their seal-of-approval. Clearly, a publication of the list of blocked sites to the general public is a bad idea, but restricting access to it to those with a reasonable reason for access would seem that there is something to hide.

    Clearly though, there are problems with ISP’s conducting censorship, especially where there is an emotional perception of youth endangerment. This has been shown in recent cases of ISP-sponsored censorship of legal-yet-unsavory content [MCI vs Epifora]. In short ISPs suck at deciding what to censor and so will any group that is thinking with their emotion before their brain. It’s very easy to say; that site is posting bad pictures, block it. Its hard to say, that site is posting bad pictures, but blocking it would block legitmiate content as well, let it be. Only the judicary has shown the ability to balance these issues effectively, so lets get them involved.

    I will point out however, that the fact they will be blocking only 12-and-under imagry doesnt make any sense, if you’re going to do it, do it right. They must provide evidence that the content is illegal, if there is not sufficient evidence to prove that the image is illegal, it cannot be blocked, there is no ‘well that image looks under 18, or under 12’ exemption. The rule of law does not allow for such subjectivity, to attach an artificial age to it would seem to limit this projects effectiveness in a way that it should not be limited; from blocking images of 12-17 year olds from outside the country! If we’re going to allow any blocking, we must allow all blocking. There can be no pick-and-choose mentality. Theres evidence, and trials. Justice, pure and simple.

    One could argue that our hate-speech laws should be protected in this way too, to ensure that Canadians do not give audience to those promoting hate in other countries; especially the US. If there is evidence of hate speech – [like godhatesfags.com], then it too should be blocked? No? If not, what is the rational of blocking child pornography but not hate speech from viewing (not to criminalize hearing, as was intended by c250).

    There are plenty of places in the world where it is legal to produce imagry in the 16 and 17 age range. This age range can be confirmed and cannot be shut down. This content is illegal in Canada, and accessing such material is illegal in Canada, and this imagry SHOULD therefore be blocked. To cop out and say 12 years only is bull. If we can verify through the rules of evidence that these images are of 16 and 17 year olds, then they too should be blocked.

    Theres no reason not to other than that cleanfeed isn’t following proper rules of evidence and is worried of making the wrong subjective call and overblocking. This to me is more troubling. There is very little subjectivity allowed in the rules-of-evidence, and if you want to have experts testify that said person could not be 18 years of age based on the evidence, so be it. But thats not for someone at cleanfeed to decide, its for a judicary to evaluate. When presented with other evidence of illegality, like the site-of-origin saying so, that too should be admissable and determinite for blocking. A 12 year old restriction is simply unacceptable – its like common carriers – it’s all or nothing. Selective enforcement of laws isn’t legal; and it cant be made legal here either.

    I hope you can work with this project to ensure it is most effective and that it guarantees the rights of Canadians are upheld. Right now the wont-someone-please-think-of-the-children crowd are saying to-hell-with-due-process if it will stop child exploitation. And this — if-it-was-your-daughter — justice system cannot be allowed to take hold. Its wrong, its driven by emotion and not evidence and its just wrong wrong wrong. Cleanfeed must at-founding have the inalienable principles of justice as it’s constitution or it simply will provide the mechanism for mallicious government, corruption and some truly scary orwellian scenarios in the future.

    Sorry about the essay, but in summary, I think the goal is noble, but the implementation has a long way to go. Lets just not lose our justice system to yet another enabling event; we’ve seen what that can lead to history, with a certain fire in germany and TWoT in the US, now child exploitation in Canada.. c’mon. Let’s learn something already. Lets do our best to protect our laws, within our legal framework. There is no excuse for breaking one law to enforce another… and as far as I can tell, this project breaks Section 36.

    To correct this, Section 36 must be ammended, it must be integrated with our legal system and our justice system must evolve with the technology. The issues of common-carrier, net neutrality, notice-and-notice[takedown with judicial oversight], blockage etc must all be evaluated, and enshrined with new legislation. Cleanfeed just isnt the right organization to be doing this type of thing, it should be a branch of the judicial system.

    K

  7. Nart V.
    1.Is there a specific set of guidelines that Cybertip uses to identify online child abuse sites/images and collect sites that are added to the blocklist sent to Canadian ISPs as part of the Project Cleanfeed Canada?
    2.What is actually collected images or web pages or web sites? Is non-web content (e.g. newsgroups) added to the block list?
    3.How are web sites stored on the blocklist? URL path to specific images? URL paths to specific web pages? Domain names? IP addresses?
    4.Is there any effort made to contact equivalent organizations to Cybertip (or law enforcement) in the country in which the offending content is hosted?
    5.Is there any effort to contact the owner of an offending website to have the offending content removed?
    6.How are ISPs blocking websites on the blocklist? By url, domain name ip address?
    7.What behavior does the user experienced if a blocked website is accessed? An error? A page telling them they’ve just tried to access illegal content?
    8.How often is the block list reviewed by cybertip to ensure accuracy?
    9.How can a blocked site request to be removed from the block list (i.e. after having removed offending content?)
    10.Is it possible to have a test research site without any illegal content added to the blocklist for research purposes?
    11. Have the ISPs received permission from the CRTC to block sites? Do they need permission from the CRTC?

  8. Continues to be unacceptable
    Thank you for investigating, Michael. But rather than allaying my fears, cybertips\\\’ responses to your queries have only stirred the pot further.

    As their responses indicated, cybertips are responsible only to those agencies and individuals who have collaborated with cybertips to introduce this ill-considered measure in the first place.

    And at the end of the day, cybertips\\\’ continued funding will depend on the number of web sites they block each year.

    Now how judiciously will they apply their bans if the benefits to cybertips of overblocking approach 100% and the risks approach zero?

  9. Lies
    Michael,

    If the anonymous people at Cyber-Tips had any intention of respecting the law and being accountable for their actions, why were none of the promised checks-and-balances put in place BEFORE the Cleanfeed project was set up?

    Or are they just patiently waiting for their critics and askers-of-inconvenient-questions to shut up and go away?

  10. Michael, please read
    I still think you are missing the point of all our arguments if you are willing to take the Cleanfeed people at their word.

    Our arguments are all based on the presumption that these folks, their motives or their system are imperfect, or will become imperfect at some point in the future.

    Your refutation of our arguments is based on your presumption (and Cleanfeeds assurances) that they, their motives and their system are all perfect and will stay that way forever.

    Ask yourself who is more likely to be proven wrong.

  11. Turnaround
    Michael,

    In 2005, you submitted a petition to the Telecommunication Policy Review Board condemning the notion that ISPs could or should block any content, including child pornography, without a proper court order.

    (PDF) [ link ]

    Why the sudden change of heart?

  12. Michael Geist says:

    a few responses
    Sorry for the long delay in responding to latest batch of comments – it has been a crazy week. A few comments in response:

    1. Project Cleanfeed is not yet operating. My understanding is that they plan to go live early next year. I absolutely agree with those that argue that Cybertip must post all of this information. In a follow-up discussion with them, I understand that to be their plan.

    2. Judicial oversight will occur before the project is launched.

    3. Posting a raw number of sites blocked will publicly identify the growth of the list that should provide some sense of the scope of the program.

    4. Blocking technologies do not block at the IP level, but are far more sophisticated, with potential blocking at the image level. That will significantly reduce collateral damage.

    5. Nart asks many good questions. I suggest contacting cybertip.ca directly for some answers. The CRTC question in particular is interesting given the questions around s. 36 of the Telecommunications Act.

  13. Wow, Michael… just wow.

    I’m loving how this blog thing can really cause a positive change.

    Do you have any specific details on point 2, regarding judicial oversight? How will this work. This is a HUGE improvement for this project.

    4. Blocking at the image level is cool, but scary too. The ability is cool, but the technological implications are kind-of disturbing as the only way to do this, is in fact, to actively monitor all the http headers, which can contain all types of pipeda covered information [get varaibles etc] that people might not like shared with a quasi-government agency.

    What is your impression on the privacy implications to this program? I can think of more than a few labour leaders might have a problem with the government watching every search string they send to google.

    Finally, you havn’t touched on the question of scope. Is cleanfeed going to be blocking any Canadian hosted sites or only extra-national sites that we can’t shut down with our criminal justice system.

    I’m glad to hear that the CRTC is involved; my initial impression is that they should deny this program operation until such time as the issue can be debated through the political process. There really should be a debate in the house about this issue, not just a debate on this blog.

    I support the goal — but the details are really really important. I’m not generally a supporter of giving up rights to gain security though, so the privacy implications, the trust issues, the political process it’s all got to be invovled… and I’m hopeful that there are good answers to all of this, but if it turns out there isn’t a satisfactory way to achieve this goal, without infinging on a bunch of other rights, then unfortunatly this program shouldn’t go forward.

    Way to protect our rights mike!

  14. Bruce Martin says:

    Time will tell
    “Way to protect our rights mike!”

    Is that sarcasm?

    I am confused (as are many like me, apparently) by Mr. Geist’s lack of cynicism regarding this latest attempt on the part of self-appointed Internet overseers to determine what is the best way to administer what is otherwise a sensible law.

    Perhaps when a few months pass, and none of Cleanfeed’s hasty promises to adopt measures of transparency materialize, then Mr. Geist will find himself once more agreeing with his more outspoken readers.

  15. It wasnt sarcasm. =P Before, we had nothing, now we’ve got judicial oversight. I don’t attribute that call to anyone but Mr Geist and the backlash on this blog.

    The program’s still far from perfect — see my prior comments, but at least its showing that this medium can work.

    We just need to keep pushing, keep demanding accountability, and if we don’t get it, well theres always psiphon if they get out of control. Fortunately, they simply don’t have all that much control over the net.

    It does bother me however, that Michael is a glass-half-full kinda guy when it comes to ‘the great firewall of Canada’. Whats up with that Mike?

  16. Bruce Martin says:

    Kevin wrote: “It wasnt sarcasm. =P Before, we had nothing, now we’ve got judicial oversight.”

    Not to split hairs, but all we’ve got now is *the promise* of judicial oversight, of an unspecified nature and according to an undetermined timeline. And this promise comes from the one party who would like judicial oversight the least.

    But I agree that it’s a good thing these people (whoever they may be) know someone is watching them.

  17. Nine months later…
    …and as of September 20, 2007 there is still no “judicial oversight”.

  18. Allready Censoring
    Cleanfeed degrades network performance and it has been used to block legitimate content already. Computer scientists have already said it can also be reverse engineered as well to provide a list of blocked sites to those with an interest in the content being blocked. In the process the content being blocked that is legitimate will likely not be realized. This is more than just dangerous- it is criminal.