{"id":1148,"date":"2021-11-22T23:12:31","date_gmt":"2021-11-22T22:12:31","guid":{"rendered":"https:\/\/monblogeur.tech\/index.php\/2021\/11\/22\/signal-loss-and-advertising-privacy-on-facebook-freedom-to-tinker\/"},"modified":"2021-11-22T23:12:31","modified_gmt":"2021-11-22T22:12:31","slug":"signal-loss-and-advertising-privacy-on-facebook-freedom-to-tinker","status":"publish","type":"post","link":"https:\/\/monblogeur.tech\/index.php\/2021\/11\/22\/signal-loss-and-advertising-privacy-on-facebook-freedom-to-tinker\/","title":{"rendered":"\u201cSignal Loss\u201d and advertising privacy on Facebook &#8211; Freedom to Tinker"},"content":{"rendered":"<div class=\"cfbc967f0983488262956e73eca9483a\" data-index=\"1\" style=\"float: none; margin:10px 0 10px 0; text-align:center;\">\n<script async src=\"https:\/\/pagead2.googlesyndication.com\/pagead\/js\/adsbygoogle.js?client=ca-pub-3859091246952232\" crossorigin=\"anonymous\"><\/script>\r\n<!-- blok -->\r\n<ins class=\"adsbygoogle\" data-ad-client=\"ca-pub-3859091246952232\" data-ad-slot=\"1334354390\"><\/ins>\r\n<script>\r\n     (adsbygoogle = window.adsbygoogle || []).push({});\r\n<\/script>\r\n\n<\/div>\n<p>November 22, 2021<br \/>  \t\t\t<a class=\"rss-topnav\" rel=\"nofollow\" href=\"https:\/\/freedom-to-tinker.com\/feed\/rss\/\">Posts<\/a>  \t\t\t<a class=\"rss-topnav\" rel=\"nofollow\" href=\"https:\/\/freedom-to-tinker.com\/comments\/feed\/\">Comments<\/a>  \t\t<br \/><a href=\"https:\/\/freedom-to-tinker.com\/\">Freedom to Tinker<\/a><br \/>Research and expert commentary on digital technologies in public life<br \/><a href=\"https:\/\/www.kyotoprize.org\/en\/en\/laureates\/andrew_chi-chih_yao\/\">The 2021 Kyoto Prize in Advanced Technology<\/a>, a major award administered by a Japanese foundation, goes to Andrew Chi-Chih Yao, a Chinese computer scientist who earned PhDs from Harvard and the University of Illinois before being a professor at MIT, Stanford, and Princeton and then becoming Dean of an important theoretical computer science education program at Tsinghua University.&nbsp; Professor Yao is a theorist, his many major important results are in \u201ccomputational complexity theory,\u201d so how did he win an international award in \u201cAdvanced Technology?\u201d&nbsp; &nbsp; Well, one of his major results led to the invention of Secure Multiparty Computation (MPC), by which two or more people can pool their data to compute a result <em>without actually disclosing their data to each other.&nbsp; <\/em>And in this article I\u2019ll explain how one present day company seems to be applying MPC to try to comply with <strong><em>privacy rules<\/em><\/strong> issued by regulators.<br \/>Facebook tracks your web browsing in order to make money delivering ads to you.&nbsp; Facebook has been under pressure from the European Union and from Apple to be less invasive of your privacy.&nbsp; For example, in 2017 the EU put out a new <a href=\"https:\/\/digital-strategy.ec.europa.eu\/en\/policies\/eprivacy-regulation\">Privacy Directive<\/a>, and in 2019 Apple\u2019s Safari browser <a href=\"https:\/\/webkit.org\/blog\/10218\/full-third-party-cookie-blocking-and-more\/\">stopped attaching cookies to third-party image requests<\/a>.&nbsp; In this article I\u2019ll discuss some indications that Facebook is beginning to adjust its advertising-tracking model so they can make money without invading your privacy quite as much.&nbsp; They are experimenting with <a href=\"https:\/\/en.wikipedia.org\/wiki\/Secure_multi-party_computation\"><em>secure multiparty computation<\/em><\/a>, a \u201c<a href=\"https:\/\/en.wikipedia.org\/wiki\/Privacy-enhancing_technologies\">privacy enhancing technology<\/a>\u201d developed in academia, to measure which ad \u201cimpressions\u201d convert to purchases <em>on the average<\/em>&#8211;but without knowing <em>which individuals <\/em>saw an ad and then made a purchase.<br \/>When you browse from one web site to another, many sites snoop on your browsing history, by <em>tracking<\/em> mechanisms such as cookies and single-pixel images (whose purpose is to track your http image-load requests).&nbsp; Much of this tracking is for the purpose of making money by targeting ads to you.&nbsp; Merchants (for example Nike) pay web sites (such as Facebook) to deliver ad views (\u201cimpressions\u201d), and Nike pays more for impressions that \u201cconvert\u201d, that is, lead to a purchase.&nbsp; So, Facebook and (independently) Nike would like to (1) <a href=\"https:\/\/www.facebook.com\/business\/help\/430291176997542?id=561906377587030\">deliver ads that are likely to convert<\/a>, and (2) measure which impressions are converting.&nbsp; Facebook wants to make more money by delivering to <em>you<\/em> the ads most likely to convert, and Nike wants to make sure it\u2019s getting its money\u2019s worth from its ad budget.<br \/>One way to do this is: when you make a purchase at the Nike online store, <a href=\"https:\/\/developers.facebook.com\/docs\/facebook-pixel\/implementation\/conversion-tracking\/\">the browser sends Facebook a copy of your nike.com shopping cart <em>and <\/em>your Facebook user ID<\/a>.&nbsp; Then Facebook looks up what Nike ads they displayed recently to that user ID; those ads <em>converted<\/em> to shoe sales, and Nike is happy to pay more for such ads.<br \/>That tracking can be a terrible invasion of privacy.&nbsp; So for years now, regulators (in California and the European Union) and browser makers (like Firefox) have been adjusting restrictions on cookies (and other kinds of tracking) to try to improve privacy.&nbsp; Facebook\u2019s internal euphemism for privacy enforcement is \u201c<a href=\"https:\/\/www.smartly.io\/blog\/brace-yourself-for-the-battle-against-signal-loss\">signal loss<\/a><a href=\"https:\/\/www.facebook.com\/business\/help\/1295064530841207?id=818859032317965\">.<\/a>\u201d&nbsp; <a href=\"https:\/\/www.smartly.io\/blog\/what-to-expect-after-the-ios-14-update\">Here\u2019s an analysis of the problem, from an advertiser\u2019s point of view<\/a> (warning: much marketing-speak!).&nbsp; The \u201csignal\u201d is the data that Facebook needs to manage its core revenue stream, advertising.&nbsp;&nbsp;<br \/>(<em>When the browser maker (Google, or Apple) is also a major advertising platform, there\u2019s an inherent conflict of interest:&nbsp; Apple\u2019s Safari restricts Facebook and Google\u2019s ad tracking more than it restricts Apple\u2019s own ad tracking, and Google-the-browser-maker delayed tightening Chrome\u2019s cookie-rules for two years because <a href=\"https:\/\/www.google.com\/adsense\/start\/\">Google-the-ad-platform<\/a> needed those cookies.)<\/em><br \/>So for years now, advertising platforms (like Facebook) have been adapting the way they intrusively track you, so they can still make money delivering relevant ads. &nbsp; For example, aside from using cookies and tracking pixels <em>inside the browser,<\/em> Nike and Facebook share things they know about you <em>outside,<\/em> like your e-mail address, phone number, and home address.&nbsp; Facebook\u2019s system for that is their \u201c<a href=\"https:\/\/www.facebook.com\/business\/help\/2041148702652965?id=818859032317965\">Conversions API<\/a>\u201d, a software interface for merchants, to measure which advertisements \u201cconvert\u201d, using server-to-server communications in the back end.<br \/>In any case, there\u2019s pressure on Facebook (and Google and other advertising platforms) to be more respectful of privacy.&nbsp; When it was just the U.S. Congress asking Zuck to testify at hearings, Facebook could perhaps laugh it off, but when entities with real enforcement power (Apple and <a href=\"https:\/\/oag.ca.gov\/privacy\/ccpa\">California<\/a> and the EU) start to insist on their users\u2019 and citizens\u2019 privacy, then Facebook might ask themselves, \u201cHow can we make money without so much privacy invasion?\u201d<br \/>Google is trying one method, called \u201c<a href=\"https:\/\/blog.google\/products\/ads-commerce\/a-more-privacy-first-web\/\">Federated Learning of Cohorts<\/a>\u201d (FLoC), but privacy advocates <a href=\"https:\/\/www.eff.org\/deeplinks\/2021\/03\/googles-floc-terrible-idea\">have severely criticized it<\/a>, for good reason: instead of sharing your entire history, FLoC labels you with a summary of your history.&nbsp; That\u2019s still a significant privacy invasion, and it may even make it <em>easier<\/em> for bad guys to track large numbers of people in harmful ways.<br \/>Is there a better way?&nbsp; Academic research on <a href=\"https:\/\/en.wikipedia.org\/wiki\/Secure_multi-party_computation\"><em>secure multiparty computation<\/em><\/a> (MPC) has shown how to measure a global property (how many Nike ads converted into sales) without identifying specific users\u2019 histories.&nbsp; In particular, with the right multiparty protocol, Nike (or Facebook) can\u2019t tell which <em>specific<\/em> purchases at nike.com resulted from ads, and they can\u2019t tell which <em>specific<\/em> ad-impressions resulted in sales, but they can measure the <em>average.<\/em>&nbsp; And that\u2019s good enough:&nbsp; good enough for Facebook to target you with ads that are more likely to convert; good enough for Nike to know that they\u2019re getting their money\u2019s worth from Facebook.<br \/>The way this kind of MPC would work is,&nbsp; the ad platform (such as Facebook) knows what ads it shows to each user.&nbsp; The merchant (such as Nike) knows which shoes it sold to each purchaser.&nbsp; They want to jointly compute the effectiveness of the ad campaign, but <em>without Facebook revealing to Nike anything about individual users, and without Nike revealing to Facebook anything about individual purchasers.&nbsp; <\/em>(<a href=\"#note1\">but see note 1 below<\/a>) So Nike would encrypt its collection of shopping carts, and Facebook would encrypt its collection of ad-impression data, and they use <a href=\"https:\/\/en.wikipedia.org\/wiki\/Homomorphic_encryption\"><em>homomorphic encryption<\/em><\/a> to compute the \u201cjoin\u201d of these relations without either one seeing the other\u2019s unencrypted data.<br \/>And indeed,&nbsp; <a href=\"https:\/\/privacytech.fb.com\/multi-party-computation\/\">Facebook claims to be adopting this method (though this explainer is very short on technical details)<\/a>.&nbsp; But Facebook has a <a href=\"https:\/\/github.com\/facebookresearch\/fbpcs\">public github repo<\/a> for their new API for advertisers, based on MPC.&nbsp; And <a href=\"https:\/\/www.facebook.com\/business\/news\/building-for-the-future\">this press release<\/a> says they\u2019re already testing their \u201cPrivate Lift Measurement\u201d with some advertisers.&nbsp;&nbsp;<br \/>Will Facebook adopt this for all advertisers?&nbsp; If they do, then I think it really will be a privacy improvement.&nbsp; It\u2019s more private than Google\u2019s solution of publicly labeling each user with a summary of their history.&nbsp; Of course, Facebook still knows <em>where on Facebook you\u2019ve been<\/em>, in every detail; and Nike still knows <em>what you browsed in their on-line store<\/em>; but nobody will know both at once.<br \/>Although MPC can <em>measure<\/em> ad conversions&#8211;whether Facebook is delivering ads that will increase shoe sales&#8211;it probably cannot <em>target<\/em> ads quite as precisely.&nbsp; That is, Facebook\u2019s machine-learning criteria to decide which ads to show you might work better if they do their super-privacy-intrusive tracking of everything you do on <em>and off<\/em> Facebook.&nbsp; By limiting their tracking to <em>on-Facebook-only, <\/em>they may find that ad impressions have a slightly lower conversion rate, so Facebook makes slightly less money.&nbsp; Time will tell whether they\u2019re willing to take that hit.<br \/>And SMP won\u2019t solve other societal problems that aren\u2019t related to privacy:&nbsp; The duopoly of Google\/Youtube and Facebook\/Instagram in online advertising, Youtube and Facebook\u2019s recommender systems pushing users towards extreme views, Youtube and Facebook trying to maximize the amount of time you waste on-line, Instagram harmful to teenage girls&#8211;none of these are about privacy, and secure multiparty computation doesn\u2019t address those problems.<br \/><a name=\"note1\">Note 1<\/a>.  Sarah Scheffler, a postdoctoral fellow at Princeton&#8217;s Center for Information Technology Policy (CITP), writes,<em> MPC&#8217;s &#8220;private&#8221; nature in these descriptions depends not only on using MPC, but also on using MPC to compute a privacy-preserving function.&nbsp; MPC could be used in the way you describe to privately compute average ad conversions, but could also be used to say, &#8220;privately&#8221; compute the list of users who are shared between Nike and Facebook or something (and I&#8217;ve heard suggestions of it being used for exactly that purpose).&nbsp; In the latter case, it&#8217;s still technically more private than Facebook and Nike comparing lists in the clear, but I don&#8217;t think it&#8217;s what most people want from &#8220;private computing&#8221;.&nbsp;<\/em><br \/>So Sarah and I took a look at <a href=\"https:\/\/github.com\/facebookresearch\/fbpcs\/blob\/main\/fbpcs\/emp_games\/lift\/calculator\/OutputMetrics.hpp\">Facebook&#8217;s open-source MPC repo<\/a>, where we see strong evidence that they are computing <em>appropriately private <\/em>functions (such as &#8220;total value of an ad campaign&#8221;).  <\/p>\n<p><textarea id=\"comment\" name=\"comment\" cols=\"45\" rows=\"8\" tabindex=\"4\" aria-required=\"true\"><\/textarea><br \/><input id=\"author\" name=\"author\" type=\"text\" value=\"\" size=\"30\" tabindex=\"1\" \/><label for=\"author\">Name<\/label> <br \/><input id=\"email\" name=\"email\" type=\"text\" value=\"\" size=\"30\" tabindex=\"2\" \/><label for=\"email\">Email<\/label> <br \/><input id=\"url\" name=\"url\" type=\"text\" value=\"\" size=\"30\" tabindex=\"3\" \/><label for=\"url\">Website<\/label><br \/><input name=\"submit\" type=\"submit\" id=\"submit\" class=\"submit\" value=\"Post Comment\" \/> <input type='hidden' name='comment_post_ID' value='16362' id='comment_post_ID' \/> <input type='hidden' name='comment_parent' id='comment_parent' value='0' \/> <br \/><input type=\"hidden\" id=\"akismet_comment_nonce\" name=\"akismet_comment_nonce\" value=\"1ca63dad9a\" \/><br \/><a href=\"#wrap\" rel=\"nofollow\">Return to top of page<\/a><br \/>Copyright &#x000A9;&nbsp;2021 &#x000B7;<a href=\"http:\/\/www.studiopress.com\/themes\/education\">Education Theme<\/a> on <a href=\"http:\/\/my.studiopress.com\/themes\/genesis\/\">Genesis Framework<\/a> &middot; <a href=\"http:\/\/wordpress.org\/\">WordPress<\/a> &middot; <a href=\"https:\/\/freedom-to-tinker.com\/wp-login.php\">Log in<\/a><\/p>\n<p><a href=\"https:\/\/freedom-to-tinker.com\/2021\/11\/22\/signal-loss-and-advertising-privacy-on-facebook\/\">source<\/a><\/p>\n<!--CusAds0-->\n<div style=\"font-size: 0px; height: 0px; line-height: 0px; margin: 0; padding: 0; clear: both;\"><\/div>","protected":false},"excerpt":{"rendered":"<p>November 22, 2021 Posts Comments Freedom to TinkerResearch and expert commentary on digital technologies in public lifeThe 2021 Kyoto Prize in Advanced Technology, a major award administered by a Japanese foundation, goes to Andrew Chi-Chih Yao, a Chinese computer scientist who earned PhDs from Harvard and the University of Illinois before being a professor at [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"googlesitekit_rrm_CAow1sXXCw:productID":"","_jetpack_memberships_contains_paid_content":false,"footnotes":"","jetpack_publicize_message":"","jetpack_publicize_feature_enabled":true,"jetpack_social_post_already_shared":false,"jetpack_social_options":{"image_generator_settings":{"template":"highway","default_image_id":0,"font":"","enabled":false},"version":2}},"categories":[1],"tags":[],"class_list":["post-1148","post","type-post","status-publish","format-standard","hentry","category-non-classe"],"jetpack_publicize_connections":[],"jetpack_featured_media_url":"","jetpack_sharing_enabled":true,"_links":{"self":[{"href":"https:\/\/monblogeur.tech\/index.php\/wp-json\/wp\/v2\/posts\/1148","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/monblogeur.tech\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/monblogeur.tech\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/monblogeur.tech\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/monblogeur.tech\/index.php\/wp-json\/wp\/v2\/comments?post=1148"}],"version-history":[{"count":0,"href":"https:\/\/monblogeur.tech\/index.php\/wp-json\/wp\/v2\/posts\/1148\/revisions"}],"wp:attachment":[{"href":"https:\/\/monblogeur.tech\/index.php\/wp-json\/wp\/v2\/media?parent=1148"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/monblogeur.tech\/index.php\/wp-json\/wp\/v2\/categories?post=1148"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/monblogeur.tech\/index.php\/wp-json\/wp\/v2\/tags?post=1148"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}