Raney v Twitter: Why the Latest Privacy Lawsuit Against a Tech Giant Would Hurt Consumers

CEO's Corner Nov 02, 2015

Tweet Image

The plaintiff’s law firm that brought data-privacy lawsuits against Amazon, Apple, Google and Netflix has struck again, warning now of a snooping menace in Twitter’s algorithms. The firm’s latest argument is a stretch – akin to accusing Microsoft of eavesdropping whenever Word autocorrects your spelling errors.

The firm Edelson PC filed a class-action lawsuit demanding that Twitter “stop its practice of systematically intercepting, reading, and altering the private messages of its users without their knowledge or consent.” Raney v. Twitter, filed last month in federal court in San Francisco, cites as examples Twitter direct messages that include links to articles that appear in the New York Times. In such cases, “Twitter’s algorithms will read through the direct message, identify the hyperlink, and replace it with its own custom link.” Pretty scary stuff. According to the plaintiff, the tactic leads to better advertising rates for Twitter. Surely something nefarious is afoot.

Privacy matters. Consumers have every right to know how their data is collected, stored, used and shared with other parties. Certainly, there are instances of companies taking liberties with personal data, using it in ways that rightfully unsettle consumers and regulators. In those cases, class-action lawsuits can serve a valuable function by policing conduct where the government will not. Raney vs. Twitter, however, is not one of those cases. Twitter isn’t listening in on your messages, or altering them. Consumers aren’t harmed. In fact, we’re all benefiting from the extremely common, safe, and well understood practices that Raney v. Twitter tries to cast as a privacy threat.  From my vantage point this lawsuit is another example of an overzealous law firm pursuing a deep-pocketed company with a flimsy, if not frivolous, case.

My reading of Twitter’s privacy policy doesn’t show any crossed lines. The company never suggests, implicitly or explicitly, that it won’t algorithmically scan private messages or change hyperlinks. But that’s splitting hairs, and beside the point.

Raney v. Twitter boils down to these questions: Can using an algorithm to substitute a hyperlink really equate with “intercepting, reading, and altering” private messages, as the plaintiffs allege? And are consumers actually hurt, their privacy violated, when Twitter’s software automatically replaces an original link with its custom one? Tellingly, the lawsuit identifies no specific harms to the named plaintiff, Wilford Raney, nor to anyone else.

None of this should surprise anyone; the law firm that brought the suit, Edelson PC, has a history of groping for novel privacy-law theories on which to sue big, successful technology companies. The firm has an “investigative team (whose) job, to put it plainly, is to find ways to sue companies,” according to a Times story published earlier this year. Its cases reap tens of millions of dollars each year for the firm. No wonder the Times dubbed the head of the firm, Jay Edelson, “if not the most hated person in Silicon Valley, very close to it.” Nor is Edelson a genuine friend to consumers.

Raney charges that Twitter alters the links within the body of direct messages. It’s true that Twitter uses its link-shortening tool algorithmically and by default. Twitter’s machines are doing exactly as they’re told: they’re substituting something that starts with http://newyorktimes and replacing it with something that starts with http://t.whatever. The “whatever” part points to the same content in the longer NYT link. When Times systems see this inbound referrer, they know that the reader came from Twitter.

Let’s say you click on the link your friend has sent to you via a direct message. When you arrive on the New York Times site, you appear to the paper as an anonymous user. At that point, NYT might drop a cookie in your browser or reference an existing cookie it dropped during one of your previous visits – standard practice consistent with the basic workings of the world-wide web. In either case, for the New York Times, you are simply “user-id 5678ABCD” – an anonymous, private reader. By reading your cookie and relating it to your past behavior on its site, the Times’ software may discern that, for example, you gravitate to stories about music. If the link in the direct message from Twitter points to another story about hip hop, the New York Times can use that insight to fine-tune its understanding of your particular musical tastes, putting you into the “music/hiphop” subcategory. That, in turn, helps the Times develop and display more relevant content for you (a review of a recent Kendrick Lamar show, for example) whenever you next visit its site.

Twitter changes the link because Twitter wants credit for sending the traffic to the newspaper. What the lawsuit ignores is how the Times and its consumers benefit.

Media companies work hard to listen to social signals for a reason: it helps them improve experiences for readers and advertisers alike. The newspaper is keenly interested in knowing how readers arrive on its site, which means the “t” at the beginning of http://t.whatever is useful to both its advertising department and its editors making coverage decisions. It helps the New York Times build a more nuanced profile of you – e.g., you’re interested in music and you’re active on Twitter – thereby displaying more relevant content. Advertisers may be interested in reaching active social media users because they’re influential and can help drive more brand engagement. Because advertisers value this signal, it’s valuable for the Times to identify anonymous ‘user-id 5678ABCD’ as someone who uses social media. Matching the right advertiser with the right consumer creates more relevant web experiences for Twitter, the Times, and you.

The signal-passing is a two-way street. Using a helpful service called Twitter Tailored Audiences, the New York Times can see its own audience segments on Twitter (e.g. active followers of music), and gain even better understandings of those segments by overlaying data collected from readers’ public tweets (e.g. identifying tweets that show a user has clicked on a story about the Coachella music festival). In addition to finding its active readers, Twitter can help the New York Times build its readership by identifying users who are similar to them, something we call ‘lookalike segments.’ This extra bit of information from Twitter helps the New York Times personalize offers and content to users who nibble at the Times’ free content but have yet to cross the paywall. That’s good for the paper, helping it turn unknown customers into paying subscribers.

Raney v. Twitter has it wrong. Twitter’s machines are just substituting a link with a slightly different format so that the companies on either end can gain helpful signals about what we like and where we’re headed. As consumers, we benefit when Twitter and the New York Times work together in a responsible and privacy-respectful way to identify the content we find most interesting. All of us – Twitter, New York Times, you, and I – get more of what we want as a result, without snooping or invasion of privacy.

The plaintiff’s law firm that brought data-privacy lawsuits against Amazon, Apple, Google and Netflix has struck again, warning now of a snooping menace in Twitter’s algorithms. The firm’s latest argument is a stretch – akin to accusing Microsoft of eavesdropping whenever Word autocorrects your spelling errors.

The firm Edelson PC filed a class-action lawsuit demanding that Twitter “stop its practice of systematically intercepting, reading, and altering the private messages of its users without their knowledge or consent.” Raney v. Twitter, filed last month in federal court in San Francisco, cites as examples Twitter direct messages that include links to articles that appear in the New York Times. In such cases, “Twitter’s algorithms will read through the direct message, identify the hyperlink, and replace it with its own custom link.” Pretty scary stuff. According to the plaintiff, the tactic leads to better advertising rates for Twitter. Surely something nefarious is afoot.

Privacy matters. Consumers have every right to know how their data is collected, stored, used and shared with other parties. Certainly, there are instances of companies taking liberties with personal data, using it in ways that rightfully unsettle consumers and regulators. In those cases, class-action lawsuits can serve a valuable function by policing conduct where the government will not. Raney vs. Twitter, however, is not one of those cases. Twitter isn’t listening in on your messages, or altering them. Consumers aren’t harmed. In fact, we’re all benefiting from the extremely common, safe, and well understood practices that Raney v. Twitter tries to cast as a privacy threat.  From my vantage point this lawsuit is another example of an overzealous law firm pursuing a deep-pocketed company with a flimsy, if not frivolous, case.

My reading of Twitter’s privacy policy doesn’t show any crossed lines. The company never suggests, implicitly or explicitly, that it won’t algorithmically scan private messages or change hyperlinks. But that’s splitting hairs, and beside the point.

Raney v. Twitter boils down to these questions: Can using an algorithm to substitute a hyperlink really equate with “intercepting, reading, and altering” private messages, as the plaintiffs allege? And are consumers actually hurt, their privacy violated, when Twitter’s software automatically replaces an original link with its custom one? Tellingly, the lawsuit identifies no specific harms to the named plaintiff, Wilford Raney, nor to anyone else.

None of this should surprise anyone; the law firm that brought the suit, Edelson PC, has a history of groping for novel privacy-law theories on which to sue big, successful technology companies. The firm has an “investigative team (whose) job, to put it plainly, is to find ways to sue companies,” according to a Times story published earlier this year. Its cases reap tens of millions of dollars each year for the firm. No wonder the Times dubbed the head of the firm, Jay Edelson, “if not the most hated person in Silicon Valley, very close to it.” Nor is Edelson a genuine friend to consumers.

Raney charges that Twitter alters the links within the body of direct messages. It’s true that Twitter uses its link-shortening tool algorithmically and by default. Twitter’s machines are doing exactly as they’re told: they’re substituting something that starts with http://newyorktimes and replacing it with something that starts with http://t.whatever. The “whatever” part points to the same content in the longer NYT link. When Times systems see this inbound referrer, they know that the reader came from Twitter.

Let’s say you click on the link your friend has sent to you via a direct message. When you arrive on the New York Times site, you appear to the paper as an anonymous user. At that point, NYT might drop a cookie in your browser or reference an existing cookie it dropped during one of your previous visits – standard practice consistent with the basic workings of the world-wide web. In either case, for the New York Times, you are simply “user-id 5678ABCD” – an anonymous, private reader. By reading your cookie and relating it to your past behavior on its site, the Times’ software may discern that, for example, you gravitate to stories about music. If the link in the direct message from Twitter points to another story about hip hop, the New York Times can use that insight to fine-tune its understanding of your particular musical tastes, putting you into the “music/hiphop” subcategory. That, in turn, helps the Times develop and display more relevant content for you (a review of a recent Kendrick Lamar show, for example) whenever you next visit its site.

Twitter changes the link because Twitter wants credit for sending the traffic to the newspaper. What the lawsuit ignores is how the Times and its consumers benefit.

Media companies work hard to listen to social signals for a reason: it helps them improve experiences for readers and advertisers alike. The newspaper is keenly interested in knowing how readers arrive on its site, which means the “t” at the beginning of http://t.whatever is useful to both its advertising department and its editors making coverage decisions. It helps the New York Times build a more nuanced profile of you – e.g., you’re interested in music and you’re active on Twitter – thereby displaying more relevant content. Advertisers may be interested in reaching active social media users because they’re influential and can help drive more brand engagement. Because advertisers value this signal, it’s valuable for the Times to identify anonymous ‘user-id 5678ABCD’ as someone who uses social media. Matching the right advertiser with the right consumer creates more relevant web experiences for Twitter, the Times, and you.

The signal-passing is a two-way street. Using a helpful service called Twitter Tailored Audiences, the New York Times can see its own audience segments on Twitter (e.g. active followers of music), and gain even better understandings of those segments by overlaying data collected from readers’ public tweets (e.g. identifying tweets that show a user has clicked on a story about the Coachella music festival). In addition to finding its active readers, Twitter can help the New York Times build its readership by identifying users who are similar to them, something we call ‘lookalike segments.’ This extra bit of information from Twitter helps the New York Times personalize offers and content to users who nibble at the Times’ free content but have yet to cross the paywall. That’s good for the paper, helping it turn unknown customers into paying subscribers.

Raney v. Twitter has it wrong. Twitter’s machines are just substituting a link with a slightly different format so that the companies on either end can gain helpful signals about what we like and where we’re headed. As consumers, we benefit when Twitter and the New York Times work together in a responsible and privacy-respectful way to identify the content we find most interesting. All of us – Twitter, New York Times, you, and I – get more of what we want as a result, without snooping or invasion of privacy.