Posts Tagged ‘Privacy Lawyers’
Behavioral Advertising and “Do Not Track”: Navigating the Privacy Minefield
Tuesday, February 28th, 2012
By Aaron Messing
The Internet is fraught with privacy-related dangers for companies. For example, Facebook’s IPO filing contains multiple references to the various privacy risks that may threaten its business model, and it seems like every day a new class action suit is filed against Facebook alleging surreptitious tracking or other breaches of privacy laws. Google has recently faced a resounding public backlash related to its new uniform privacy policy, to the extent that 36 state attorney generals are considering filing suit. New privacy legislation and regulatory activities have been proposed, with the Federal Trade Commission (FTC) taking an active role in enforcing compliance with the various privacy laws. The real game changer, however, might be the renewed popularity of “Do Not Track”, which threatens to upend the existing business models of online publishers and advertisers. “Do Not Track” is a proposal which would enable users to opt out of tracking by websites they do not visit, including analytics services, advertising networks, and social platforms.
To understand the genesis of “Do Not Track” it is important to understand what online tracking is and how it works. If you visit any website supported by advertising (as well as many that are not), a number of tracking objects may be placed on your device. These online tracking technologies take many forms, including HTTP cookies, web beacons (clear GIFs), local shared objects or flash cookies, HTML5 cookies, browser history sniffers and browser fingerprinting. What they all have in common is that they use tracking technology to observe web users’ interests, including content consumed, ads clicked, and other search keywords and conversions to track online movements, and build an online behavior profiles that are used to determine which ads are selected when a particular webpage is accessed. Collectively, these are known as behavioral targeting or advertising. Tracking technologies are also used for other purposes in addition to behavioral targeting, including site analytics, advertising metrics and reporting, and capping the frequency with which individual ads are displayed to users.
The focus on behavioral advertising by advertisers and ecommerce merchants stems from its effectiveness. Studies have found that behavioral advertising increases the click through rate by as much as 670% when compared with non-targeted advertising. Accordingly, behavioral advertising can bring in an average of 2.68 more revenue than of non-targeted advertising.
If behavioral advertising provides benefits such as increased relevance and usefulness to both advertisers and consumers, how has it become so controversial? Traditionally, advertisers have avoided collecting personally identifiable information (PII), preferring anonymous tracking data. However, new analytic tools and algorithms make it possible to combine “anonymous” information to create detailed profiles that can be associated with a particular computer or person. Formerly anonymous information can be re-identified, and companies are taking advantage in order to deliver increasingly targeted ads. Some of those practices have led to renewed privacy concerns. For example, recently Target was able to identify that a teenager was pregnant – before her father had any idea. It seems that Target has identified certain patterns in expecting mothers, and assigns shoppers a “pregnancy prediction score.” Apparently, the father was livid when his high-school age daughter was repeatedly targeted with various maternity items, only to later find out that, well, Target knew more about his daughter than he did (at least in that regard). Needless to say, some PII is more sensitive than others, but it is almost always alarming when you don’t know what others know about you.
Ultimately, most users find it a little creepy when they find out that Facebook tracks your web browsing activity through their “Like” button, or that detailed profiles of their browsing history exist that could be associated with them. According to a recent Gallup poll, 61% of individuals polled felt the privacy intrusion presented by tracking was not worth the free access to content. 67% said that advertisers should not be able to match ads to specific interests based upon websites visited.
The wild west of internet tracking may soon be coming to a close. The FTC has issued its recommendations for Do Not Track, which they recommend be instituted as a browser based mechanism through which consumers could make persistent choices to signal whether or not they want to be tracked or receive targeted advertising. However, you shouldn’t wait for an FTC compliance notice to start rethinking your privacy practices.
It goes without saying that companies are required to follow the existing privacy laws. However, it is important to not only speak with a privacy lawyer to ensure compliance with existing privacy laws and regulations (the FTC compliance division also monitors whether companies comply with posted privacy policies and terms of service) but also to ensure that your tracking and analytics are done in an non-creepy, non-intrusive manner that is clearly communicated to your customers and enables them to opt-in, and gives them an opportunity to opt out at their discretion. Your respect for your consumers’ privacy concerns will reap long-term benefits beyond anything that surreptitious tracking could ever accomplish.
Have You Really Thought About the Practices You Preach?
Thursday, March 17th, 2011Your Privacy Policy Could Have Serious Legal Implications
How many times have you seen website terms of use or privacy policies saying something to the effect, “We use industry standard best-practice technology to guarantee your sensitive financial transactions are 100% safe and secure?” When you publish these types of statements, you potentially expose your business to deceptive and/or unfair practices claims by attorneys general, state and federal regulators, and private plaintiffs, particularly if there is a data breach involving sensitive information. From a business perspective you may not like the more watered down version, “While we take reasonable measures to try to protect your sensitive information, we cannot guarantee that your information will be completely secure, etc…” However, industry standards are made to be broken by the nefarious crews who make it their work to steal financial account access numbers, as well as other sensitive, information. If you think that you provide the panacea to all online risk, speak up! You may have discovered the golden goose. Until then, think about publishing more accurate, responsible information for your users and to mitigate your business risk. Besides, being accurate creates user confidence, and these things can be worded in ways to build trust in your brand.
Protecting data applies when it is in transit and at rest. That means that after you receive the data through an encrypted connection, there are risks related to its storage; if, and when, it is unencrypted and used. Interestingly, the recent HBGary Federal hack against a well-known information security firm demonstrated that even those charged with the task of protecting information are susceptible. In creating your public facing policy, have you focused on security after only the transmission stage?
About that encrypted transmission, many times these industry standards utilize Transport Layer Security (TLS) and its predecessor Secure Sockets Layer (SSL) technology. You know these, they create the HTTPS standard. We’re often advised to look for the “HTTPS” in the URL heading, or the lock icon in our browser. In my travels I am astonished to learn that some people think these technologies are infallible. So, once that happens, our connection is secure and invincible, right? Well…maybe.
While the detailed workings of TLS and SSL are way beyond this article (and certainly beyond my ability to fully appreciate) it is interesting to note that researchers have found potential vulnerabilities with SSL, or at least with the supporting browser and trusted authorities concepts necessary for its use in typical online transactions. This is not to say that TLS and SSL are not safe. Quite the contrary, the encryption technology provides good protection for sensitive online transactions and should definitely be used. However, they must be configured correctly, the Certificate Authority (CA) must act appropriately, and the client (user) machine must not be compromised. The security and confidentiality sought through the use of SSL depends upon not only the encryption algorithm, but also the browser and the trust aspect inherent in public key cryptography.
Regarding the encryption itself, while some proclaim that they use “industry standard” technology, they might actually not be using it. SSL version 2.0 was known to have several security vulnerabilities. The Payment Card Industry Digital Security Standard (PCI DSS) does not recognize SSL Version 2.0 as secure. Only Version 3.0 or other later TLS standards may be considered.
Browsers by default can be loaded to trust numerous CA’s. CA’s are entrusted to determine that the site that it claims to be, is actually that site as claimed. In the past researchers had found that known vulnerable certificates had not been revoked by some CA’s, and theoretical or actual “collisions” where a man-in-the-middle assumes the trusted identity could happen.
Would it surprise you that according to some analysis, some certificates might still support SSL Version 2.0? According to one researcher, as of July 2010 only about 38% of sites using SSL are configured correctly, and 32% contain a previously exposed renegotiation vulnerability. Other researchers exposed approximately 24 possible exploits (of varying criticality) involving man-in-the-middle attacks on SSL when used in browsers.
Most recently in February 2011 Trusteer reported on some nasty malware they named OddJob. OddJob targets online banking customers. According to Trusteer, OddJob does not reside on the client and thus avoids detection by typical anti-malware software. A fresh copy of OddJob is fetched from a command and control server during a session. OddJob hijacks a session token ID, and reportedly allows the hacker to, essentially, ride-along in the background with the user’s session. Of most concern, OddJob allows the hackers to stay logged in to one’s account even after the user purports to log-out; thus, maximizing the potential for undetected (or later detected) fraud. Significantly, client side (user-based) malware presents possible risk, some of which may be beyond the online website’s control.
So, if we presume that no technology will be absolutely 100% safe and secure, and if the right bad-guys want to target someone or something, why the need to tell users something that is not necessarily accurate?
This is only one example of good practices in vetting what you are actually doing to see how it really measures-up, and how your public facing policies may seem accurate, when they really are not. This article focuses on one aspect of security, but the same types of issues arise in privacy as well. Why expose your business to more regulatory risk if there is a breach? Even if you employed good practices and did your best to try to protect the information, false or misleading information in your public facing terms and policies can come back to haunt you.
Appointing experienced information governance individuals or teams, or using outside resources, can help you identify the disconnects and gaps between what exists, and what you say exists.