November 26, 2007
The Future of Reputation: Gossip, Rumor, and Privacy on the Internet: Ackerman's Review of Daniel Solove's Newest Book on Privacy and the Internet
My slightly incented [FN1] take on Dan Solove's most recent book: The Future of Reputation: Gossip, Rumor, and Privacy on the Internet.
by Ethan Ackerman
Unfortunately, I have to advise against beginning to read this book late on a Friday night on a crowded subway car full of antics-heavy party-goers if you also have a cellphone camera in your pocket. Why? It's not the locale or the company, as Dan Solove begins his compelling second exploration of privacy on a crowded subway too - on the freshly dog-poop-messed floor of a Korean subway car, to be exact. No, it's because readers that are most of the way through the book can appreciate the situation's potential much better. Fortunately for my fellow Friday night riders (especially the guy who kept trying to play the jug on an empty bottle of malt liquor) I was already into Chapter Six that night. More than 130 pages in, I'd already heard tales of funnier, sadder, and more embarrassing moments recorded for public posterity. More importantly, I'd also been exposed to a significant chunk of Solove's excellent thinking on the nuanced impact on individual privacy that comes from a massively participatory Internet - where everyone can be a content contributor, where anonymity and obscurity are likely fleeting but consequences may be lifelong, and where many users' expectations and norms are still trying to adapt to previous technologies like cameras and newsprint.
It is this aspect of Solove's book - the deep AND wide thinking about an individual's interaction with the modern Internet - that moves the book out of the one-point-rigorous-analysis of an academic article and the semi-random anecdotal topicality of a blog post and into the category of critical (in the must-read sense) literature. Where Solove's previous work tackled the pressing but somewhat solve-able problems that arose from individuals losing control of their personal information to government and commercial entities, this book tackles individuals' loss of access and control of their information at the hands of other individuals - and, increasingly, by their own hand on blogs, social networking and image sharing sites of their own.
As Solove points out, this book is much more ambiguous and less proscriptive in its conclusions than his prior work. Reading the book is the best way to understand why Solove doesn't just say "do more 'x'," or "make 'y' illegal on the Internet," or "tech people about how to do 'z'," but Solove's reasoned reluctance to throw out lots of solutions (and the nuance of the few ideas he does suggest) seem to be the main sticking point with reviewers so far. While having my own suspicions and opinions on what might and might not (and what definitely does and does not) work about modern privacy protections, I have to agree with Solove on one of his main premises - it is equal parts impossible and vital to avoid over-censorship or under-protection when trying to simultaneously protect individual autonomy, speech and privacy. I'd have to agree with what I think Solove's ultimate aim is here - informing people and getting them to think more about privacy themselves. To put words in Solove's mouth, if everyone is more informed and thinks about these issues themselves, not only will any ultimate solutions probably be better, but they will also perhaps be moot, as more people will have chosen the non-problematic action in the first place.
Chapter Three may prove particularly interesting for some of the marketing and economics-analysis readers of this blog. Despite the fact that it looks like a well-written lay discussion of just what makes communication different on the Internet (it even innocently starts out with anecdotes about the Washingtonienne scandal), Chapter Three is an excellent stealth lesson in communications theory, information theory, and even an economics lesson about reducing transactions costs and capturing the benefits and harms of previous externalities. Solove could have titled this chapter "the long tail phenomenon of online talk." Though the chapter never says so in so many words, it discusses the impact of markedly reduced transaction costs of sharing AND recalling information. It uses network theory to explain how the connectedness of an information sharer still largely dictates information flow, and it also explores how the Internet has increased both the positive and negative externalities associated with information exchange.
My biggest beef is with a topic that pops up only briefly, once at the end of Chapter Two and again briefly in the conclusion: what might the future of privacy look like? More specifically, where are important trends taking us and in what ways may we consider privacy that we don't now? Solove's work dedicates page after page to the values and desires and social benefits that form our notions of privacy, and equally numerous pages to developments that alter and possibly enhance or diminish our privacy. But how about at least exploring alternate visions of what that may look like as time progresses? (Talk about reviewers wanting to have their cake and have it fed to them too.) The last two paragraphs of Chapter Two tease the reader with the idea that privacy may not even matter to future generations, a plausible idea that finds much support in the preceding pages. Yet, this meme is basically dropped until a brief resurrection in Chapter Eight, where it serves (in "Do people want privacy anymore") as a foil for Solove's convincing later argument that yes, people want privacy and they have a nuanced, non-binary view of privacy that changes, but does not necessarily diminish, with each technological change. How about a few pages on that dys/u-topian future where privacy doesn't matter - the utopian one where social acceptance and factual accuracy negate any need to 'hide' embarrassing facts and early learning mistakes, or the dystopian one where our warts and peccadilloes are now universally surveilled and broadcast, but still have sanctions and social stigmas associated with them?
[FN1] If there is only one thing for an author worse than a poor book review, it is probably no book reviews. While this problem wouldn't apply to The Future of Reputation, Dan Solove did a rather innovative thing to insure against it anyway. He offered free review copies to any mainstream blogger who would author a review, regardless of content. This review is in response to that generous offer.
November 21, 2007
Search Redirection Tool Could Be Trespass to Chattels--Burgess v. EForce
By Eric Goldman
Burgess v. EForce Media, Inc., 2007 WL 3355369 (W.D.N.C. Nov. 9, 2007)
Every now and then a consumer goes on a me-vs.-the-world bender and decides to unilaterally save society by suing everyone in sight. Burgess' anger over unwanted advertising may have sparked such a campaign. His previous appearance on the blog involved his pro se lawsuit against American Express and many other major brand names for unwanted pop-up ads. In that ruling, the court intimated that advertisers could be liable for contributory trespass to chattels.
In this companion action, Burgess sued a number of defendants for spam. The court rejects his CAN-SPAM claim for lack of standing (he doesn't qualify for the limited private causes of action).
Burgess also sued for the installation of search redirection client software, claiming it was a privacy invasion, trespass to chattels, and "illegal conduct." The defendants first tries to dismiss the claims as preempted by CAN-SPAM, but CAN-SPAM's preemption clause does not apply to generally applicable laws like privacy invasions and trespass to chattels. Nevertheless, the magistrate report (approved by the judge) dismisses the privacy invasion claim for failure to state a claim, saying:
While the undersigned shares in plaintiff's frustration with the internet and the unconscionable applications that interfere with one's use and enjoyment of technology--and at times display offensive websites--frustration of purpose is not an invasion of privacy. Further, the undersigned cannot find any North Carolina case recognizing a cause of action for invasion of privacy based on computer viruses that redirect internet searches or inquiries, or any cases that would suggest that similar such conduct in other fields would support such a claim.
The "illegal conduct" claim was also dismissed.
On the other hand, building on Burgess prior ruling in state court, this court refuses to dismiss the trespass to chattels claim. Citing to Sotelo and others, the court says that Burgess' "pro se pleadings are not a model of clarity but nevertheless suffice to state a claim for trespass to chattels. He sufficiently alleges actual possession of his computer and 'unauthorized, unlawful interference' with his use of this personal property." So the Sotelo precedent marches on, even though this court (as with the prior Burgess court) doesn't acknowledge Hamidi, Mummagraphics or the other cases that would put these expansive trespass to chattels rulings in serious doubt.
As a result, Burgess' case lives to see another day. I'm sure we haven't heard the last from him!
November 19, 2007
Roommates.com Amici Curiae Briefs
By Eric Goldman
Amici curiae briefs were filed in the Ninth Circuit Roommates.com appeal by:
* a variety of Internet companies such as Google, eBay and Amazon plus non-profit organizations such as the EFF
* various news organizations
* the ACLU. Roommates.com has taken the ACLU brief as unfriendly and wrote a reply brief.
Other documents in the case:
* The Fair Housing Councils' request to brief Batzel. Roommates.com's opposition. The Ninth Circuit denied the Councils' request on Nov. 6.
* The Ninth Circuit order granting the en banc hearing
* Fair Housing Councils' reply to the EFF et al amicus brief
* EFF et al amicus brief supporting a rehearing en banc
* Fair Housing Council's response to Roommates.com's request for an en banc rehearing
* Roommates.com's En Banc Request
* The original Ninth Circuit opinion
* My blog post on the Ninth Circuit opinion
November 18, 2007
Online Word of Mouth and its Implications for Trademark Law
By Eric Goldman
I've posted my latest academic work, Online Word of Mouth and its Implications for Trademark Law, to the web. This is a chapter in a book edited by Graeme Dinwoodie and Mark Janis containing contributions from leading trademark scholars from around the world.
Online word of mouth describes chatter about marketplace offerings that historically has taken place in the workplace hallways or in casual interactions between family/friends but is now electronically mediated. In the first part of the chapter, I argue that online word of mouth warrants exceptionalism from trademark law because it represents an unprecedented type of demand-shaping content that is both outside the control of trademark owners but has a potential global reach. In the second part of the chapter, I show how existing trademark law improperly inhibits online word of mouth. Specifically, this chapter ties into the "trademark use in commerce" debate by arguing that "commercial referential trademark uses"--the mechanism that flips online word of mouth into trademark law--should be categorically immune from trademark scrutiny.
I'm particularly excited about this chapter because it ties together a number of themes that I've been addressing on this blog for the past (nearly) 3 years. I know a lot of you are interested in online trademark issues, and this chapter is a "succinct" (well, <10,000 words) distillation of my analysis and arguments on these topics. So if you want to see my thoughts about online trademark law sketched out in a single integrated discussion, read this article. Plus, it makes for excellent plane/train/car reading on the way to/from your Thanksgiving destination!
This Chapter discusses the emergence of online word of mouth, the process by which consumers disseminate their views about marketplace goods and services. Due to online word of mouth, consumers have an unprecedented ability to influence the brand perceptions of other consumers. Unfortunately, these effects have put doctrinal pressure on trademark law, leading to judicial interpretations that inhibit the flow of online word of mouth and may damage the efficacy of marketplace mechanisms. This Chapter will explore how trademark law should be interpreted to preserve the flow of socially beneficial online word of mouth.
November 17, 2007
Google and Yahoo Sued for Hosting Content That Allegedly Infringes Trademarks--(ISC)2 v. Degraphenreed
By Eric Goldman
International Information Systems Security Certifications Consortium v. Degraphenreed, 2:07 CV 1195 (S.D. Ohio complaint filed Nov. 16, 2007)
International Information Systems Security Certifications Consortium ("ISC2") offers a professional designation entitled "Certified Information Systems Security Professional," or "CISSP" for short, that individuals can earn by meeting the published requirements. The Consortium has a federally registered certification mark (#2045256) for the term "CISSP." The complaint alleges that Degraphenreed was once a registered Certified Information Systems Security Professional but he failed to satisfy the continuing standards. As a result, the complaint alleges that Degraphenreed now describes himself as a "Chief Information Security Systems Practitioner," also abbreviated as "CISSP," thereby continuing to claim CISSP status without meeting the ISC2's standards.
These allegations appear to support trademark infringement and false advertising claims, although interestingly I can't find any examples of Degraphenreed's usage of the term "Chief Information Security Systems Practitioner." (I got zero results in both Google and Yahoo searching for the term "Chief Information Security Systems Practitioner."). ISC2 also alleged trademark dilution but that should be a non-starter because I doubt CISSP will qualify as widely recognized among the general consuming public.
The most interesting aspect of this case is that ISC2 also sued Google and Yahoo for trademark infringement for hosting content that contained Degraphenreed's impermissible CISSP usage. Specifically, the complaint alleges that Google hosted six blogs that contained the CISSP mark (at least 2 of which contained the term in the blog title), and that Google refused to take down these blogs after the plaintiff's notice. The complaint also alleges that Yahoo hosted 5 Yahoo Groups referencing CISSP and a Flickr account containing ISC2's CISSP logo, and that after plaintiff's notice Yahoo only removed one group and left everything else up. The complaint claims direct (not contributory) trademark based on these allegations.
From my outsider's perspective, it looks like a significant tactical error to bring Google and Yahoo into this lawsuit for at least four reasons:
1) The plaintiff's theories of trademark liability against Google and Yahoo are untested and lack any useful precedent. In fact, to date we really don't have an exemplar lawsuit discussing the liability of a service provider for hosting trademark-infringing content, and I can't think of a case where a service provider has been held liable a trademark infringer for hosting user content. This claim reminds me a little of the Jews for Jesus v. Google Blogspot lawsuit from Dec. 2005 (which ultimately settled irresolutely), where Jews for Jesus complained about a third level domain/blog title selected by a blog user. When that lawsuit was filed, I speculated about some of the possible theories of liability and defenses, but the law was murky at best. So in this case, suing Google and Yahoo makes a relatively straightforward case much more complex and expensive.
2) Often, individual defendants in these types of cases don't hire top-flight IP defense lawyers....but Google and Yahoo most assuredly will. As a result, ISC2 has ensured that some very skilled attorneys will line up on the defense to break every aspect of its case.
3) I couldn't investigate everything, but what I saw of Degraphenreed's activities on Google and Yahoo didn't look immediately problematic. For example, some of the blogs really lack any substance at all (see, e.g., here), but they don't look like splogs. If anything, it looked like ISC2 may be trying to shut down some griping. For example, two of the Yahoo groups are entitled "cissp-clueless" and "cissp-censorship," and the cissp-censorship group is a restricted access group with only three members. It's not clear how this group could possibly contribute to a trademark infringement claim. Instead, it looks like ISC2 might be overreaching, perhaps to shut down some unwanted commentary, and this may increase the judge's sensitivities to the public interests at stake here.
4) The plaintiff can get all of the relief it needs just by suing Degraphenreed. If the plaintiff wins that lawsuit, they can get an order forcing Degraphenreed to remove the infringing material. Further, I imagine that Google and Yahoo would happily take down any content that a court has adjudged infringing.
Please email me if you have any thoughts about why ISC2 decided to go after Google and Yahoo (let me know if I can post your comments). For now, I'm classifying it as a blunder. It will be interesting to see how aggressively Google and Yahoo respond to this lawsuit.
UPDATE: I've posted some updates here.
November 16, 2007
The Victorian Internet
By Eric Goldman
As you might infer, I'm not at the cutting-edge of reading books. I don't read that many books in general; and I tend to read books well after everyone is done talking about them. But I enjoyed this book so much, and it has aged so well over the past decade, that I thought it was worth a shoutout. One side bonus: given that the book is no longer a hot release, you should be able to get this book used for well less than $10.
The book discusses the history of the telegraph. The book explains the technologies preceding the telegraph, the battles between the inventors of the telegraph, the telegraph's role in spawning new technological innovations (and creating enormous wealth for some of those folks) and the ways that the telegraph did--and did not--change society.
Its thesis is that many phenomena we associate with a global electronic network first occurred in the 19th century, not the 20th, which has made our celebration of the Internet's novelty (a topic at its zenith in 1998 when the book was published) ahistorical. The book thoroughly delivers on this thesis. One particular anecdote really hammered this point home. The book talks about a telegraph-mediated "online wedding" that first occurred...before 1848. (Indeed, the book "Wired Love" was published in 1879 and an article "The Dangers of Wired Love" ran in 1886). Yet, numerous newspaper articles from the mid-1990s marveled at Internet-mediated weddings as if they were completely unprecedented.
More generally, the book broadly makes the case that some things never change. For example, the book describes the arms race between telegraph companies establishing pricing schemes to curb attempts to send more information at a lower cost, just to have telegraph senders coming up with new gaming strategies. The book discussed the paranoia of major institutions in response to telegraphy, including governments that sought to control the use of cryptography in telegraphy and newspapers that assumed that the telegraph would destroy their business. (In the latter case, the newspapers adapted and thrived in response to telegraphy). The book also described how the telegraph contributed to feelings of information overload.
The book ends on a bittersweet note. It observes that people thought that the borderless telegraph communication network would contribute to world peace by breaking down barriers to communication. It didn't. If anything, the telegraph played an important role in 19th century imperialism and contributed to some of the bloodiest wars in history. Similarly, 150 years later, many similarly romanticize how the Internet can make the world a better place. Perhaps the Internet is truly different from the telegraph in this respect, or perhaps, we are just ahistorically proclaiming the latest technology innovation as our savior. As the book says, "That the telegraph was so widely seen as a panacea is perhaps understandable. The fact that we are still making the same mistake today is less so."
From my perspective, the only thing "missing" from this pithy and efficient book was a more thorough discussion of how lawmakers reacted to the rise of the telegraph. I would like to know more about how 19th century regulators coped with--or, more likely, freaked out about--the technological assumptions changed by the telegraph.
It seems safe to assume that some legislators misunderstood the technological underpinnings of telegraphy. The book gives numerous examples of how people didn't understand that the telegraph sent only electronic signals and wasn't a teleportation technology, such as the story of a woman in 1870 who sought to "telegraph" sauerkraut to her son. Again, some things never change; in 2003, a member of the House of Lords had a similar misunderstanding about spam. [the exact quote: "Will the Minister explain how it is that an inedible tinned food can become an unsolicited email, bearing in mind that some of us wish to be protected from having an email?"]
In this vein, the book offered one possible explanation for Sen. Stevens' explanation that the Internet is a "series of tubes." [the exact quote from Wired: "the internet is not something you just dump something on. It's not a truck. It's a series of tubes. And if you don't understand those tubes can be filled and if they are filled, when you put your message in, it gets in line and its going to be delayed by anyone that puts into that tube enormous amounts of material, enormous amounts of material."] Many telegraph operators built out a network of pneumatic tubes to move messages over short distances because this was quicker and more accurate for those messages. So perhaps Sen. Stevens was thinking of the telegraph when he referred to the "series of tubes."
The book was published before Western Union sent its last telegram. At the time I knew this represented the end of an important era, but after reading this book I better understand the significance of that event. Some day, someone will send the very last TCP/IP enabled http message...an event that will also probably pass with a whimper, not a bang.
November 13, 2007
Geolocation and A Bordered Cyberspace
By Eric Goldman
I recently gave a talk on the general theme of the future of e-commerce, and I was allowed to take the topic in any direction. I decided to talk a little about the propagation of geolocation technology and its consequences for a borderless Internet. My notes from the talk:
A constant problem in Cyberlaw: the difficulties of authenticating users for age and geography. With respect to geography, in the mid-1990s, there was a strong belief that cyberspace was borderless. Examples:
* John Perry Barlow's 1996 Declaration of the Independence of Cyberspace
* 1997: ALA v. Pataki, where a state anti-Internet porn law (a baby CDA) was struck down as violating the dormant commerce clause. In that case, Judge Preska said: "Geography is a virtually meaningless construct on the Internet."
But there are ways to restore geographic borders to the purportedly borderless Internet:
1) Ask users to self-report. Users may want to self-report geography, especially in the e-commerce context where they want physical goods delivered or need to report their address to authorize a credit card purchase. But the law could force online actors to compel users to self-report geography and then act on the reported information. Examples:
* LICRA v. Yahoo. The French court envisioned that Yahoo could do 90% effective geographic authentication through a combination of IP address analysis and user self-reporting if Yahoo popped up windows asking users to self-report before being allowed to access the website.
* Alaska SB 140, an anti-adware law. To combat pop-up ads, the statute requires software vendors to display pop-up windows asking users to self-report geography.
A world with compelled requests for user self-reporting of geography would be a pop-up filled world constantly asking "where are you now? where are you now?" [see the analogous Verizon ad campaign] This makes user self-reporting undesirable, in addition to being unreliable.
2) IP address analysis. IP addresses are allocated on an International scheme. Yahoo used this scheme to display local ads, a fact noted in the LICRA court. IP address analysis can be more regional; for example, Google does geo-targeting on a more granular basis. Ex: if I search for "mercedes" in Google, I get local Mercedes dealers in the Bay Area. But IP address analysis is incomplete/imperfect.
So if the only geographic authentication tools were IP address analysis or user self-reporting, the Internet would remain more borderless than bordered. However...
3) Geolocation technology. In the future, Internet access devices will be coupled with GPS technology that will automatically report user geography. For example, many mobile phones already have GPS technology in them, and consumers use other mobile devices (e.g., Blackberries) that have geolocation technology. Inevitably, the boundaries between computers and these geolocated mobile access devices will dissolve, meaning that Internet access devices will be geolocated and will automatically self-report user geography as part of interacting with other online actors.
A geolocated Internet will have some benefits. Most obviously, ads can be geographically targeted in ways that can help consumers (i.e., a driver searching for gas can get ads from nearby gas stations). It will also enable other localized content where that matters (weather, directions, location of friends).
But a geolocated Internet will also enable governments to force online actors to "honor" the geographic information. Thus, states could legitimately enact state-specific laws and require online actors to customize their offerings for state residents. Governments could also use the geolocation information to created walled environments, including more highly filtered/screened content. We've already seen this in China and some other countries. In these situations, Internet users will have very different Internet experiences based on their geography. Thus, a geolocated Internet should contribute to the demise the Internet utopianism. Instead of bringing people together over a borderless network, a geolocated world reenables borders that will keep us further apart.
November 10, 2007
Google Resists Subpoena for Keyword Ad Purchases--Connor Sport Court v. Google
By Eric Goldman
Connor Sport Court International, Inc. v. Google Inc., CV-06-3066 PHX JAT // CV 07-80252 (N.D. Cal. motion to compel filed Oct. 31, 2007)
This summer, I reported on trademark litigation between Connor Sport Court and Rhino Court. The parties had settled the lawsuit, but then Connor complained that Rhino violated the settlement by buying keyword advertising triggered to Connor's trademarks. Connor then submitted a discovery request to Google seeking records of other people who had bought Connor's trademarks as keywords. As I noted at the time, the requested information had significant competitive value, and Google's delivery of the information could prompt a lot of other similar discovery requests to Google.
Initially, Google seemed inclined to give Connor the data it asked for, but apparently Google changed its mind. Instead, Google has refused to turn over any data related to third party purchases and didn't turn over much related to Rhino. Connor apparently still believes the requested information is worth pursuing, because it has now filed a motion to compel Google to comply with its discovery request.
Google might take the opportunity to clarify its policies regarding the disclosure of keyword ad purchases. Connor's brief claims that Google provided Rhino with information about a third party's ad purchase, including the ad copy, the maximum cost-per-click bid, the number of clicks and impressions, the average ad position and more. Is Google handing out this information merely based on a subpoena, or is Google going to make it harder for litigants to get access to this data? According to the filing, the hearing is scheduled for Dec. 7 at 9 am in San Jose.
November 06, 2007
Ripoff Report Gets Mixed 230 Ruling--Children of America v. Magedson
By Eric Goldman
Children of America, Inc. v. Magedson, CV 2007-003720 (Arizona Superior Ct. Oct. 24, 2007)
In a ruling reminiscent of Hy Cite v. badbusinessbureau, an Arizona trial court dismissed a plaintiff's defamation claims against Ripoff Report based on third party content, but didn't dismiss the claims that alleged Ripoff Report wrote the content. This is the correct application of 47 USC 230 on a motion to dismiss, but now the plaintiff has to show that the defendant authored the content in question, and the court says ominously "Should Plaintiff be unable to prove that Defendants authored any actionable statement, the remainder of this case will likely be susceptible to resolution by way of summary judgment."
While the court reaches the right result, the court gets goofy when it says "published case law construing [47 USC 230] is surprisingly scant." It's only surprising if you don't do thorough research. Heck, the Ripoff Report has generated at least a half-dozen reported opinions itself.
One final point of interest. The court says confidently "Nor can [defendants] face liability for their actions in promoting the site, organizing its content, making the contents more accessible on search engines or soliciting contributions of content." While the court doesn't cite the Roommates.com opinion, it seems to be going out of its way to reject Kozinski's opinion.
November 02, 2007
Chandler on Regulating Search Engine Bias
By Eric Goldman
I find the topic of search engine bias especially fascinating because it seems to bring out strong pro-regulatory sympathies in many people. Just about every academic (other than me) who has written on the topic has sought ways to counteract any bias effects, typically through some form of regulatory intervention of the search industry.
The latest pro-regulatory article on this topic is Jennifer Chandler, "A Right to Reach an Audience: An Approach to Intermediary Bias on the Internet", Hofstra Law Review, 2008. Jennifer's article provides a great starting point if you want an overview of the issues and a summary of the pro-regulatory arguments. Among other highlights, the article explores the linkages between the search engine bias issue and the network neutrality debate and acknowledges the First Amendment doctrines that protect online intermediaries from regulation of their freedoms of expression.
The article anticipates some of my critiques of its arguments (made here), so I won't replay that discussion here. But the following sentence from Jennifer's article really caught my attention:
We have an opportunity with these novel information intermediaries to craft new approaches to the intermediary function that enable us to keep as much of the Internet's free speech benefits as possible in a context where we are hopefully not bound by "the way it has always been."
As this makes clear, the article embraces both search engine utopianism and exceptionalism. Starting from those premises, the article makes commendable arguments. But if those premises are wrong (and I think they are), then the resulting conclusions are destined to be wrong as well.
Some of the thorniest problems of communications law and policy were supposed to have been solved by the Internet. The issue of who can speak, or access the means of speech, was said to have been solved by the arrival of ubiquitous, relatively cheap access to the Internet. The problem of media concentration was supposed to have been solved now that so many more speakers could contribute. While the Internet has undoubtedly assisted with these problems, new gatekeepers have arisen, and that their actions are not necessarily supportive of a healthy, pluralistic communications environment. While the problem of access to the means of speech seems to have been greatly alleviated by the Internet, the chokepoint has now shifted downstream to a class of intermediaries that select and filter information en route to listeners. Examples of this class of selection intermediaries include search engines, software filters, Internet Service Providers (ISPs) that block or filter content, and spam blocklists.
It is true that we have long been surrounded by too much information, and we have relied on various intermediaries to assist us in finding and choosing information. Why, then, is the role of selection intermediaries on the Internet worthy of comment? In my view, the Internet offers an opportunity for us to craft new approaches to the selection intermediary function in a way that enables us to keep as much of the speech freedom engendered by the Internet as possible. There is a danger that by reflexively drawing analogies to familiar old selection intermediaries, such as libraries, we will tolerate selection criteria that erode the freedom of speech made possible on the Internet.
In the age of the Internet, a complete theory of communication rights must explicitly address the effects of selection intermediaries and recognize as protected each of the steps involved in the communicative relationship between speaker and listener. This includes not only the right to speak and the right to hear (which are already recognized forms of free speech rights), but also the right to reach an audience free from the influence of extraneous criteria of discrimination imposed by selection intermediaries. If selection intermediaries block or discriminate against a speaker on grounds that listeners would not have selected, that speaker's ability to speak freely has been undermined. The paper makes a case for the recognition of this right. It also considers whether government regulations to give effect to this right could be imposed without violating the free speech rights of selection intermediaries.