Spoiler: Avoid blocking only by client-side! DISCLAIMER: All data and information provided in this article are for informational purposes only. The main goal is to increase security awareness, teach about information security, countermeasures and give readers information on how to implement a safe and functional system (in our case, a website). If you plan to use the information for illegal purposes, please leave this website now. I cannot be held responsible for any misuse of the given information. I do not use this method to avoid paying and neither should you. The word “hacking” usually has a strong meaning and most people would think of gaining root access to the website I am referring to in the title or insight into its database. In this case, however, I want to demonstrate how you can exploit a specific error in its design and . gain access to a feature that was initially blocked Starting from the beginning, I enjoy reading scientific articles on the website. There is however a certain aspect that bothered me. , for more (along with other perks) you need to subscribe to a paid for “all-access digital program”. www.technologyreview.com You can only read three free articles per month Every time I read a story, a banner would appear telling me that I have read 1/3, 2/3 or 3/3 of the stories I was entitled to for that month and when I tried to open a new article, after I’d read 3 stories in a month, it prompted me to subscribe in order to have access to it. So I started thinking of how this functionality worked and how it could be bypassed. One could circumvent that perhaps by (or ) and reading three articles from each one, but making a lot of accounts constantly changing IP addresses I don’t consider that to be a reasonable solution. I started wondering how the stories I was reading were being counted. My was that original hypothesis every time I clicked on a URL (I was signed in to my account), a request was made to the server, the number of articles I had read in a month was checked for being less than 3, the article ID was kept (so I could revisit that article) and then I would get the corresponding response from the server (the article itself or a message that I needed to subscribe for more). This seemed pretty logical, but I decided to as well. I started thinking like a programmer and with my beloved searched for certain variable names and bingo! take a look at the source code (Ctrl-U) Ctrl-F Here is what I found with some reverse engineering logic. First I spotted the following snippet of HTML code that indicated the message that I was to be shown according to the : number of articles I’d read <!-- Reading Meter --><div class=" "><span show="meter.loggedIn.1"><span class="meter__line"> <span text="meter.usageWord"></span> </span> meter You've read of three free articles this month. After that, I searched for the word “meter” in the source code, since it is everywhere in the above snippet. What I came up with was a script assigning a JSON object type to a variable named “serverData” (!!!). Here is what a part of it looked like: = "user":{"id":1111111,"username":"John D","email":"myemail@gmail.com","status":"active","rss_token":"st45pew5rX2UQmFafadsgajtP92a","livefyre_id":"uid_11111111","notifications_unread":0,"profile":{"status":"public","display_name_from":"username","name":"John D","url":"/profile/john-d/","first_name":"John","last_name":"D"}, <script> serverData "meter":{"month":"2018-01","count":1,"ids":[600000],"hits":[{"id":600000,"date":"2018-01-01T15:11:32-04:00"}]},"email_notifications":....</script> Below that, there were some more scripts with a link to JavaScript files, so I started opening them too: <script src=" 2222222" defer></script> https://cdn.technologyreview.com/js/example.js?v=2 From searching, in what seemed to be a lot of lines of JavaScript, I managed to isolate some “revealing” lines of code (again using Ctrl-F and searching for keywords). Here are some of my findings: function P(e,t){e.func.call(e.context,t,e. )} count++ function w(){if(!y())return!1; ,....loggedOut:e.ids.length+1>o.meter.allowedMax?]} o.meter={usage:i.length>h.allowed? {if(o.meter={usage:e.ids.length>h.allowed?h.allowed:e.ids.length,allowed:h.allowed,allowedMax:h.allowedMax,visible:!0,open:!0,paywall:!1,usageWord:"",allowedWord:"",loggedIn:{1:!1,2:!1,3:!1,"3b":!1} h.allowed:i.length,allowed:h.allowed,allowedMax:h.allowedMax,...,loggedIn:{1:!1,2:!1,3:!1,"3b":!1},loggedOut:{"1a":!1,"1b":!1}},o.meter.loggedIn[o.meter.usage.toString()]=!0,null (S=new Date).getMonth()...t(window).trigger("mittr:meterBlock",[{ ,... e&&(b=11 reason:"User is above the allowed view limit" It was obvious that both the increment of the counter and the inspection of whether the read articles were less than three, were made at client-side in JavaScript!! (it actually turned out you didn’t even need to be logged in) The way someone could overcome the 3 articles per month restriction was now in front of me. I just had to block JavaScript on the website! It may be possible to benefit from even more mistakes in the website's design from this kind of server-side validation, but after discovering this basic flaw, I sent an email to MIT Technology Review reporting my findings. Summing up, there are two basic mistakes in this implementation: a) The validation of the number of articles that the user has read in one month is made at client-side, so by disabling JS, one has unlimited online access. b) The counter is incremented at client-side, so by disabling JS, the number of read articles remains the same. So even if the validation was made at server side we could still have access to more than three articles. I also have to mention that by blocking JavaScript new cookies were not created . If you delete the old ones there is no way to access data concerning the user’s access, with the current implementation at least. So another way to bypass the paywall is to delete or block cookies. Someone might think that this kind of vulnerability is not that serious, but that is not quite true in my opinion. First of all, it shows and indicates that a lack of basic “best practices” more mistakes may be found in the website's implementation. Another important aspect is . In this case, for example, to gain access to the feature of . Depending on the website’s this might be an (that is acquired by offering unlimited online access as a service). the feature’s importance “unlimited online access to articles” you need to purchase a subscription business plan important source of revenue The lesson to be taught here, as simple and fundamental as it might sound to people with a lot of experience on web developing, is: avoid blocking only by client-side! A great hacking challenge that also taught me this, is “HackGame3, by ChauRocks” , https://hackgame.chaurocks.com (solutions/hints at https://github.com/KAUTH/HackGame3-solutions ). This article is a real-life example of one of the tricks you need to think of to proceed to the next levels of the game. UPDATE: After writing this article . Moreover, in some cases, you can , something that was not allowed before. I came across many other websites whose paid prime program can be circumvented with the same method (blocking JavaScript) browse having your add-blocker on The sites affected that I have found so far are: MIT Technology Review , Business Insider , The New York Times , The Washington Post , The Economist , The Boston Globe All the above websites have been informed about the flaw. Learned something? Hold down the 👏 to support and help others find this article. Thanks for reading!! Follow me on Twitter @konpap1996
