Mr R. Pitofsky Chairman Federal Trade Commission 601 Pennsylavia Av NW Washington DC 20580 February 22, 1999 Dear Sir, This letter is to ask the Commission to consider action it might take to prevent Intel's Processor Serial Number (PSN) from severely damaging consumer privacy and consequently stunting the growth of ecommerce. Most of the media coverage of this issue has incorrectly reported that "Intel disabled the feature" or that "the number will be off by default." Both these statements are false. Intel did not change the chip at all; they merely decided to change the "control utility" software they will provide to OEMs (PC manufacturers), which the OEMs may or may not use. The page of questions and answers on the PSN published on Intel's own web site on or before 2/3 stated that ``While the processor serial number is activated in the chip, the default control utility setting will turn the feature to "OFF." The utility then allows the user to choose whether to enable the processor serial number feature...'' Nothing there suggested that this "default off" setting would be anything less than universal. Yet in a 2/18 Reuters story an Intel official admitted that OEMs will be shipping PCs with the PSN on (contrary to the statement above) in units destined for the workplace. So in addition to the statements attributed by the media to Intel, Intel's own directly published statements have not accurately reflected the reality of their position. We request that the Commission consider whether its Section 5 authority regarding false claims and deceptive practices should be brought to bear on Intel. Beyond the question of deception is the issue of the harm that will be caused by Intel's actions, whether misrepresented or not. While the case against the PSN has been stated in many places (such as http://www.bigbrotherinside.com) the following argument has been formulated along lines familiar from other actions taken by the Commission. In summary we believe that: 1) The PSN is likely to cause substantial harm to consumer privacy and consequently reduce consumers' participation in ecommerce. 2) This harm will not be easily be avoidable by consumers. 3) The harm will not be outweighed by countervailing benefits. The following paragraphs expand these three points in turn. 1) The PSN is likely to cause substantial harm to consumer privacy and consequently reduce consumers' participation in ecommerce. This assertion is based on the following subassertions. (a) The PSN will be become a de facto standard Global User Identifier (GUID). (b) The GUID will be used by companies in information practices that are unfair. (c) Such practices will become known to consumers, some of whom will avoid participation in ecommerce because they apprehend that their privacy is at risk by doing so. (a) The PSN is destined to become a de facto standard Global User Identifier (GUID) for the Internet, much as the Social Security Number became the GUID for financial transactions. Intel's stated intention to add the PSN to their other chips, plus their near-monopoly market share mean that the PSN feature would be present on the majority of PCs in a few short years. Although other uniquely identifying numbers have been available on computer hardware, none has had the ubiquity and attractiveness of the PSN. Intel has listed copyright protection as one of the advantages of the PSN, and software publishers are certain to adopt it for the consumer software market. (In the business software market, similar mechanisms have been available on expensive workstations for years.) Any piece of software that requires the PSN for copyright protection could obviously also use it for other purposes. (b) The GUID will be used by companies in information practices that are unfair. The history of cookies has shown that browser manufacturers and web sites have a mutual commercial interest in tracking and targeting consumers using mechanisms that are turned on by default. Several companies have already set up schemes to bypass the limitation of cookies that each site gets a different cookie, allow the sharing of information about visitors. Having a PSN provided by the browser to web sites would allow these extra efforts to be bypassed. Given that the browser market is a duopoly where both duopolists have a significant consumer ecommerce operations, those manufacturers have a compelling incentive to make the PSN available to their own sites and to partners. Beyond browsers, other software such as mail and chat programs might choose to disclose the PSN, overtly or covertly. (Intel have already indicated one of the intended uses is to exclude "rogue users" from chat rooms.) In the current legal environment where web sites (except those targeted at children) are not required to abide by any code of fair information practice, and where information on online behavior is a valuable and salable, the PSN will inevitably be used in unfair, privacy-invasive practices. (c) Such practices will become known to consumers, some of whom will avoid participation in ecommerce because they apprehend that their privacy is at risk by doing so. Surveys such as the 1998 Harris/Business Week poll indicate that fear for privacy is the number one reason consumers give for not going online (ahead of price and usability). The PSN will move reality much closer to these consumers' worst fears. The considerable media attention and public discussion of Intel's announcement reflects the high level of concern over the state of online privacy. 2) This harm will not be easily be avoidable by consumers. This point follows from each of two assertions. (i) Many consumers will not know they are using a PSN-enabled machine, or will not understand the implications of the PSN for privacy. (ii) Many will be compelled to disclose their PSN. The most extreme example of compulsion will occur in machines in the workplace. Intel have listed asset management as an intended use of the PSN. To support this, organizational buyers would order PCs from the manufacturers with the PSN permanently enabled in the BIOS. It could be argued that employees should have no expectation of privacy at the workplace, but lunchtime private usage actually raises the level of Internet activity, and many consumers find the cost of establishing an Internet connection at home prohibitive, so this group will represent a substantial number of ecommerce participants. Furthermore, business routinely dispose of PCs to employees, schools and others for personal use, and few consumers know how to reconfigure a BIOS. Even on PCs where use of the PSN is ostensibly optional and opt-in, experience shows that consumers will be coerced into submitting to the tracking mechanism. Microsoft for example demands cookies as a precondition to access to large amounts of technical information, some essential to performing key tasks. Other sites require cookies or registration as a condition of entry to the site, or as a condition of purchase, or a condition of using software. This may be as essential as the operating system or personal finance software. The putative choice is often illusory. Returning to point (i), many consumers will not know they are using a PSN-enabled machine, or will not understand the implications of the PSN for privacy. Intel has suggested a small taskbar icon as a visual indication, but this decision is not even in their hands, and manufacturers are unlikely to voluntarily place a warning label on their products that explains the privacy implications of a PSN. 3) The harm will not be outweighed by countervailing benefits. Intel claims various benefits for the PSN, mostly in improving security. However as documented on http://www.bigbrotherinside.com mentioned above, several leading technical experts have stated that as a security mechanism the PSN is too weak to be very useful. In areas such as asset management and copyright protection, other mechanisms are already in use that do not depend on a PSN. This completes our argument that the PSN is likely to cause substantial harm which will not be easily be avoidable by consumers, and will not be outweighed by countervailing benefits. As you may know, privacy groups have consistently called on Intel to permanently disable the feature since the day it was announced. Some have asked Intel to recall all product shipped with the feature. We ask the Commission to consider what action it might take to reduce the harms to consumer privacy and ecommerce identified here, including any means to compel the company to disable the feature and order a recall, whether directly or through PC manufacturers. We also request that Commission consider whether its Section 5 authority regarding false claims and deceptive practices should be brought to bear on Intel. Finally, we request a meeting with the appropriate staff of the Commission to discuss this matter. This letter is not a formal petition and complaint to the Commission, but such a plea might follow at a later date. The Commission's consideration of this letter and any response will be very much appreciated. Very respectfully Jeffrey Chester The Center for Media Education 2120 L Street, NW Suite 200 Washington, DC 20037 Russ Smith Consumer.net PO Box 44232 Washington, DC 20026-4232 Marc Rotenberg Electronic Privacy Information Center 666 Pennsylvania Av SE Washington DC 20003 Jason Catlett Junkbusters Corp. P.O. Box 7034 Green Brook NJ 08812 Simon Davies Privacy International London (c/o EPIC in USA) Evan Hendricks Privacy Times P.O. Box 21501 Washington, D.C. 20009 Beth Givens Privacy Rights Clearinghouse 1717 Kettner Ave. Suite 105 San Diego, CA 92101 Robert Bulmash Private Citizen, Inc. PO Box 233 Naperville, IL 60566 (Correspondence may be addressed to EPIC, who will summarize and distribute responses, or to all signatories if desired.)