Law in the Internet Society





A Proposal for the Regulation of Internet Data Mining by Private Entities

-- By BrettJohnson - 11 Nov 2009


Internet data mining touches on all three components of privacy discussed in class: (1) secrecy, (2) anonymity, and (3) autonomy, which in turn are intertwined with personal freedom. Freedom and autonomy are desirable and true freedom requires meaningful choice. The challenge and desire then is to determine and implement the legal paradigm that provides people with the best meaningful choice and freedom with respect to their privacy.

Privacy issues arise in part because of reasonable expectations of privacy for certain activities.

What does this mean? It appears to be tautology.

For example, when a woman posts photos to a Facebook profile and that site states that people can control who can see their profile and photos she has a reasonable expectation of privacy when those privacy options are exercised. Conversely, if a man uses an online username “BOYSRCH” along with his photograph, which is accessible by anyone, he should not be surprised if such is used against him by his employer with a policy that prohibits homosexuality. This type of activity does not have a reasonable expectation of privacy and no legislative reform is necessary to protect such publically revealed information.

Maybe or maybe not. The discharge here is occurring not because of a policy against homosexuality, but a policy against disclosure of homosexuality, the so-called "Don't ask, don't tell" nonsense. It's not a very good example of the principle, whatever it is, that you're supposedly illustrating.

Aside from the reasonable expectations of privacy, people should be entitled to maintain their privacy as a natural right in certain matters that they do not choose to disclose. For example, the identity of products that a person purchases over the internet should be protectable as a natural right—irrespective of any disclosures that would eliminate a reasonable expectation of privacy—should that person choose privacy.

What process of distinction separates the "natural right" situations, whatever that phrase is supposed to mean here, which is unexplained, from those in which you would apply the equally-misadopted "reasonable expectation" idea? You don't give any underlying logic for your conclusions, as usual, you just pronounce them, as though assertion were a substitute for analysis.


Various privacy options, e.g., absolutely prohibiting data mining, unlimited data mining, opt-in, nakedness, and opt out systems have some difficulties either in theory and/or in practical administration. The current unlimited and unknown to many internet data mining violates fundamental human rights. See generally Planned Parenthood of Southeastern Pennsylvania v. Casey, 112 S.Ct. 2791, 2486 (U.S. 1992) (“Throughout this century, this Court also has held that the fundamental right of privacy protects citizens against governmental intrusion [into certain areas].”).

Whatever reproductive privacy has to do with privacy, government intrusion has little to do, given our general constitutional dogma, with activities by non-government parties. Once again, an analytic issue is assumed away rather than dealt with.,\

A per se outright ban on all data mining is also problematic. This is a more difficult issue than the former but with the ideal that true freedom means individual choice, one must recognize that people should be free to allow monitoring and use of information if after being fully informed they subjectively perceive that such monitoring benefits them more than it costs them.

I would have thought the initial issue was whether data-miners possess constitutional rights to learn, cogitate and teach what they have learned and inferred.

An “opt-out” system is undesirable for the basic reason that people are simply provided with too many complex (often probably intentionally so) adhesion form-contracts to be expected to carefully read and understand such, resulting in effectively no choice and nearly unlimited data mining in such a system.

Finally nakedness impermissibly negates personal choice to maintain privacy (although it would reduce incentives to gather the information).

This is no more than a brief and fairly insensitive paraphrase of points I initially made. No thinking here yet.


An opt-in system provides the best internet privacy option. As a default starting position, people are entitled to privacy and they must take action before such privacy is relinquished. Of course there are legitimate reasons that persons would want to know certain information about a person—such as the credit history and ability of a prospective tenant to pay rent. Another example might include the political (and possibly even personal activities) of a public figure such as a U.S. Senator. In the former case, however, the tenant is choosing to disclose certain personal information to the landlord through contract. In the latter case the U.S. Senator is voluntarily relinquishing certain privacy rights by becoming a public figure (the Senator’s formerly personal information becomes relevant to the public because the Senator’s behavior could directly affect the public).

There is not a bright line where one person’s right to information overcomes another’s right to privacy. However, society should error on the side of privacy with privacy rights only being overcome where there is a compelling showing of need for the information or the person giving up her privacy rights has made a knowing decision to do so under the circumstances. An opt-in system for data mining provides the best opportunity for implementation of the potentially competing goals of obtaining relevant information while protecting privacy rights.


Legislation should be adopted wherein the default rule provides that without consent in the form of “opting-in,” information gathered about a person over the internet may only be used as necessary to provide the service requested. The information gathered could not be disclosed or sold and it would need to be deleted within a reasonable amount of time. For example, if a person placed an order from all information about the purchaser, including her personal information such as name, address, etc. and the product purchased, web pages visited, etc. would need to be deleted from Wal-Mart’s database within a reasonable time after the product is received by the customer (discussing Wal-Mart’s current use of personal information). Other entities, such as the Google search engine would not be able to store or disclose the information (private browsing available from Firefox). In the context of banking, information such as expenditures would need to be retained for record-keeping purposes but kept confidential and not used for purposes other than record-keeping and such information could not be shared with other entities or other departments within the same institution (such as where investment banks are allowed to merge with commercial depository banks after repeal of the Glass-Steagall Act).

The legislation would, however, allow those private entities to mine data (purchase the ability to store, use, and sell the information) if after being fully informed a person believed that it was in her best interest to sell that information and opted-in. Some people may desire to have special offers sent to them for future purchases of similar products while others may be persuaded by discounted prices or even cash payments for the information. The opt-in choice would perhaps require that to opt-in the person would be redirected to a federally maintained website that provided in understandable and brutally descriptive terms (drafted as part of the legislation) what the information could be collected, used for, by whom, and potential consequences thereof. There could also be an option, each time referred to the “opt-in” website to register a single time to preclude all companies from making future offers to mine data from that IP address.

Why spend all these words on something that needs no further explanation, unless because you have nothing of your own to say? Would it not have been sufficient to state that European data protection principles are preferable to North American ones? And would that not have led predictably to some discussion of the real politics here, which you have not so far mentioned and might be therefore thought to have tuned out completely while I was discussing them in class?


While not free of concerns, an opt-in system provides the best choice, freedom, and overall autonomy for individuals in society.

An improvement over previous efforts only to the extent that there are no glaring inaccuracies. There are no new points made, no analysis of any legal, political, or technical issues that I didn't myself discuss, no new information gathered, and nothing new proposed or investigated. Low effort, low commitment, slight improvement.

You are entitled to restrict access to your paper if you want to. But we all derive immense benefit from reading one another's work, and I hope you won't feel the need unless the subject matter is personal and its disclosure would be harmful or undesirable. To restrict access to your paper simply delete the "#" on the next line:

# * Set ALLOWTOPICVIEW = TWikiAdminGroup, BrettJohnson

Note: TWiki has strict formatting rules. Make sure you preserve the three spaces, asterisk, and extra space at the beginning of that line. If you wish to give access to any other users simply add them to the comma separated list


I added this comment box. Feel free to delete it if you didn't want one; just find the percentage-sign COMMENT percentage-sign text in the editing screen and delete it.

The paper reads well. In terms of suggestions, I have only a few. First, you should link to Steven Wu's essay in discussing that some people choose aggregation in return for customized services. You might also look at Dana Delger's essay in that regard. I tend to agree that an opt-in system is the best approach, for the reasons you outline: those who wish to sign-up for data aggregation should be allowed to do so, and free-for-all monitoring is highly problematic. If I were to suggest revision to the essay, I would think saying more about just what sort of information people must be given when deciding to opt-in or not would be helpful. You don't want the opt-in option to be like the Windows Vista User Access Control, but you need it to also not overload the user with info (as you suggest). Maybe on the federal site you mention, you could also have a sort of opt-in wiki? Where people can post and discuss the results of their opting-in and out. The initial opt-in/out button gives essential, basic info (just a few sentences), and the link takes you to the federal page and a wiki where you can learn more? Whatever the optimal opt-in model, I think saying more about ideas for it would be helpful. Otherwise, it looks pretty good.

-- BrianS - 20 Nov 2009


I have a two suggestions, one on style and two on substance.

First, rolling links into words instead of having them stand alone in the text would make the paper flow more smoothly in some instances. To place a link you can use this syntax:

So if I wanted "Google" to hyperlink to I would use the above syntax with:
LINK TEXT = Google
So for example, if you wanted to put a link into this sentience: All three components of privacy and in particular autonomy are intertwined with personal freedom.

You could change it to this: All three components of privacy and in particular autonomy are intertwined with personal freedom.

Second, although I agree that opt-in is better than the current system, I find this draft confusing as to why you think so. In section II you introduce five different privacy options: "absolutely prohibiting data mining, unlimited and unregulated data mining, opt-in, nakedness, and opt out systems." You then proceed to reject four out of the five and thus accept the fifth one as "the best chance" for privacy. But you've said nothing about the merits of opt-in at all. Its like if I had 4 rocks and threw them at 5 targets and concluded that the one target I did not hit was unable to be hit by rocks. I think your proposal for legislation might be bolstered by spending the space on what is good about opt-in instead of what is bad about the other options.

Finally, I think that if you have the space you may want to try to address some of the counter arguments to your proposed legislation. If Google cannot store any data about searches, are we prohibiting Google from knowing something that happened to it? Does this have First Amendment implications? Will your proposed legislation prevent me from keeping a log of the visitors who come to my personal website? Another objection is that it prevents Google from protecting itself. Servers can be attacked through many different methods. A DDoS attack is an attack where many computers send millions of requests for information (for example, sending a search request to Google) to the same server with the goal of overloading the server and taking it offline. These attacks often target sites like Google and Twitter. One way of surviving the attack is to identify the computers sending the requests and stop accepting requests from those computers. Your legislation would prohibit Google from identifying who is attacking their computers and to take measures to prevent the attack from succeeding.

-- JustinColannino - 22 Nov 2009


Thank you for your helpful comments. I will put some thought into them and try to address them in a revised version of the essay.

-- BrettJohnson - 22 Nov 2009

Brett, You may end up addressing my comment based on some of the previous comments, but my suggestion is slightly different than previous comments so I'll leave it for you as well. I agree with your position that an opt-in system would be better than many other options, but I think that you dismiss the possibilities afforded by an opt-out system a bit too quickly. If legislation required that users first be provided a clear explanation of what data could and could not be sold, saved, etc. and were then provided with the option to opt out, wouldn’t the opt-out system be functionally equivalent (or at least very similar) to your proposed opt-in system? I think the difference between the two systems may come down, in part, to whether the default rule is to favor privacy or “business interests” (at least as that term refers to the businesses gathering data). While I agree that opt-in is a better option, I think that your argument against opt-out would be strengthened by acknowledging that it is inferior at least in part because of a decision that maintaining privacy and control over one's personal information should be the default (in other words, that control of one's personal information is a value in itself).

-- HeatherStevenson - 23 Nov 2009

  • I don't think Justin's points are to be dealt with as space allows. Collecting information and making judgments about it is constitutionally-protected learning. You can no more make it required that I get people's permission to learn about them than that I get their permission to learn mathematics. What I know about the senior Senator from New York or the President of Columbia University or the guy who rents my house in the country is not something I can be commanded by anyone to forget, or refrain from using to make inferences that are either useful or beautiful. Any other result vitiates the very idea of free expression. If that proposition is true, your argument fails completely. If that proposition is false, where and how is it false?



Webs Webs

r15 - 07 Sep 2011 - 00:44:08 - IanSullivan
This site is powered by the TWiki collaboration platform.
All material on this collaboration platform is the property of the contributing authors.
All material marked as authored by Eben Moglen is available under the license terms CC-BY-SA version 4.
Syndicate this site RSSATOM