The US Federal Trade Commission has released a new report on consumer privacy that tries to cover the waterfront on concerns Americans have about everything from tracking their online and mobile habits to the anonymous video analytics technology being heavily marketed these days by Intel and a host of smaller companies.
The FTC report is 112 pages and looks to be pretty heavy going, but a very quick skim got me to the stuff relevant to this sector – the issue of pattern detection technology being used to capture images of consumers, analyze them, and serve ads or information accordingly.
The report – Protecting Consumer Privacy In An Era Of Rapid Change – has a specific section on AVA.
DATA ENHANCEMENT CASE STUDY: FACIAL RECOGNITION SOFTWARE
Facial recognition technology enables the identification of an individual based on his or her distinct facial characteristics. While this technology has been used in experiments for over thirty years, until recently it remained costly and limited under real world conditions.
However, steady improvements in the technology combined with increased computing power have shifted this technology out of the realm of science fiction and into the marketplace. As costs have decreased and accuracy improved, facial recognition software has been incorporated into a variety of commercial products. Today it can be found in online social networks and photo management software, where it is used to facilitate photo-organizing, and in mobile apps where it is used to enhance gaming.
This surge in the deployment of facial recognition technology will likely boost the desire of companies to use data enhancement by offering yet another means to compile and link information about an individual gathered through disparate transactions and contexts. For instance, social networks such as Facebook and LinkedIn, as well as websites like Yelp and Amazon, all encourage users to upload profile photos and make these photos publicly available. As a result, vast amounts of facial data, often linked with real names and geographic locations, have been made publicly available.
A recent paper from researchers at Carnegie Mellon University illustrated how they were able to combine readily available facial recognition software with data mining algorithms and statistical re-identification techniques to determine in many cases an individual’s name, location, interests, and even the first five digits of the individual’s Social Security number, starting with only the individual’s picture.
Companies could easily replicate these results. Today, retailers use facial detection software in digital signs to analyze the age and gender of viewers and deliver targeted advertisements. Facial detection does not uniquely identify an individual. Instead, it detects human faces and determines gender and approximate age range.
In the future, digital signs and kiosks placed in supermarkets, transit stations, and college campuses could capture images of viewers and, through the use of facial recognition software, match those faces to online identities, and return advertisements based on the websites specific individuals have visited or the publicly available information contained in their social media profiles.
Retailers could also implement loyalty programs, ask users to associate a photo with the account, then use the combined data to link the consumer to other online accounts or their in-store actions. This would enable the retailer to glean information about the consumer’s purchase habits, interests, and even movements, which could be used to offer discounts on particular products or otherwise market to the consumer.
Summary and Recommendations
The ability of facial recognition technology to identify consumers based solely on a photograph, create linkages between the offline and online world, and compile highly detailed dossiers of information, makes it especially important for companies using this technology to implement privacy by design concepts and robust choice and transparency policies. Such practices should include reducing the amount of time consumer information is retained, adopting reasonable security measures, and disclosing to consumers that the facial data they supply may be used to link them to information from third parties or publicly available sources.
For example, if a digital sign uses data enhancement to deliver targeted advertisements to viewers, it should immediately delete the data after the consumer has walked away. Likewise, if a kiosk is used to invite shoppers to register for a store loyalty program, the shopper should be informed that the photo taken by the kiosk camera and associated with the account may be combined with other data to market discounts and offers to the shopper. If a company received the data from other sources, it should disclose the sources to the consumer.
This is pretty consistent with what the Digital Signage Federation has recommended with respect to best practises and disclosure, and also what Intel has worked on with the Ontario Privacy Commissioner, which developed the Privacy By Design principles referenced in the FTC report.