6 ms·
How Carrier IQ was wrongly accused of keylogging
- rhizome 15y agoIsn't it a little early to say "wrongly," while the companies are still playing hot potato?
- anigbrowl 15y agoThat's never stopped him before.
- brlewis 15y agoQuote: It's true that carriers already know what URLs you're visiting when you use their network Is that true for https?
- podperson 15y agoThey know the domain but not the rest of the url or the data being exchanged. But hey, he's a political science writer so he clearly knows more about all this than any of us.
- sp332 15y agoThey only know which server you're visiting, so they can route the packets properly. But the headers are encrypted, so they shouldn't get the whole URL. https://en.wikipedia.org/wiki/Https#Network_layers https://en.wikipedia.org/wiki/Https#Network_layers
- brlewis 15y agoI'm familiar with the normal workings of https. What I'd like to know is if there's anything special about cellphone browsers such that the carriers act as MITM (like with Kindle Fire). What I can't imagine, though, is any way they could MITM the URL and not be able to MITM the rest of the HTTP request and response.
- lurker17 15y agoObviously the browser can MITM anything. The browser renders the page content for you and translates your input into network traffic.
- rmc 15y agoAll ISP can do loads of packet sniffing. Remember FireSheep? So in theory they could intercept every non ssl request and see what sites your going to. But things like that might be illegal, and expensive. Kindle Fire is a different kettle of fish and is almost certainly not used by most ISPs.
- marshray 15y agoNon-smartphone browsers will often use a bastardization of https that does allow the carriers to perform MitM interception. This should* not be the case with smartphones however, particularly when connected to wifi and not even transferring data via the carrier. *The carrier may have planted an SSL trusted CA cert on your phone before giving it to you.
- Dylan16807 15y agoNobody ever accused carrier IQ of hacking the software onto the phones all by themselves. Tracking all URLs is half as bad as tracking keys. And forgive me if I don't find claims that it doesn't record the keystroke data that it's sent anyway in a customizable app very reassuring.
- freehunter 15y agoHow is tracking all URLs half as bad as a keylogger? You understand that any ISP can see a log of which customer requested which website, right? They don't need CarrierIQ to see that, it's an integral part of the function of being an ISP.
- stevelosh 15y agoYou understand that when using https an ISP only ever sees the hostname, not the rest of the URL, right?
- freehunter 15y agoYes, I understand this. I don't see your point. Are URLs sensitive information? Hiding a URL is security through obscurity, I don't expect an outrage over obscurity being revealed. At least not from anyone who understands how security works.
- bad_user 15y agoThis is not about security, this is about respecting your privacy. There's a difference between: www.facebook.com ... and ... www.facebook.com/rebecca.some.name
- freehunter 15y agoBut doesn't the carrier see "1-555-123-4567" "Rebecca S. Name" "facebook.com" "123 Fake St, Springfield Ohiyamaude" "Visa XXXX-XXXX-XXXX-XXXX" ? Wouldn't they be able to put 2 and 2 (and 2 and 2 and 2 and 2) together to know who the traffic is coming from anyway?
- ghshephard 15y agoTurns out they aren't wrongly accused of keylogging. Carrier IQ, likely in partnership with either the Handset Vendor and/or the carrier were responsible for the storing of personal information. Here is a pretty good quotation from their director of marketing: ""We're as surprised as anybody to see all that information flowing," Andrew Coward, Carrier IQ's director of marketing, told CNNMoney in an interview. "It raises a lot of questions for the industry -- and not [only] for Carrier IQ.""
- marshray 15y agoThey're just trying to throw their customers (the carriers) under the bus at this point. They really don't want to shoulder these oncoming class actions alone.
- earl 15y agoThis is stupid. If they don't want to be accused of being a keylogger and/or rootkit, they should (1) not allow keypress events to be sent to their software; (2) force carriers to not hide their executable and show users what information they are collecting on their phones; (3) keep a record of information, again to be shown to users on demand from the program installed on their phones, that has been uploaded; (4) manage to answer simple questions, like (i) could you forward keystrokes, sms, ssh passwords, ssl passwords, etc, (ii) can your software contact your servers and download new control instructions, (iii) has this ever happened, (iv) under what circumstances would it happen (5) easily give users the ability to shut down the flow of information and remove the program, including in all future updates Really, this isn't rocket science.
- jonmc12 15y ago(6) not advertise ability to "Capture button presses" - http://www.wired.com/images_blogs/threatlevel/2011/12/ExperienceManager.datasheet-1.pdf http://www.wired.com/images_blogs/threatlevel/2011/12/Experi... (7) not include "user interaction" and "key presses" within the core claims of their technology patent http://www.faqs.org/patents/app/20110106942 http://www.faqs.org/patents/app/20110106942 ..just might be construed as intent regardless of evidence.
- rmc 15y agoI agree with your sentiment and most of what you say. But a lot of the general public are not good at computers. Remember the "facebook login" incident? If you tell these people that there a magic programme that runs all the time then they will always turn that off. As a result carriers will haves basically no information.
- marshray 15y agoYou mean people will be able to make informed choices about their privacy?
- rmc 15y agoNo, lots of people do not know about how to make informed choices about lots of things about tech. You right know how to use computers, but there are loads of people who Google "www.Facebook.com" they do not know about computers
- podperson 15y agoPlease change title to "How Carrier IQ's PR Department is able to post articles on cnet".
- corin_ 15y agoWhile I agree with you in terms of content, based on how they've handled the rest of this mess I find it almost hard to believe they would be capable of looking up a cnet editor's email address, yet alone pursuading them to write anything.
- marshray 15y agoThey are not having any trouble with being in contact with the press at this point. What seems more likely is that they were able to feed the reporter enough material from their side of the story that the reporter used mostly that to meet his deadline. Of course, the reporter may write a story from the opposite perspective Monday.
- goombastic 15y agoWell actually, they probably have everyone's email address and they probably wrote it for him... using key presses on his mobile.
- tptacek 15y ago"Step 1: get a well-known consultant from VSR to reverse engineer your code gratis and publicly post that the allegations made about your software appear groundless". You probably don't even need a step 2.
- kevinalexbrown 15y agoNot keylogging does not absolve Carrier IQ. This article doesn't exactly say that, but it gives the impression that it's all one big misunderstanding from an overzealous "newly minted" security researcher. That people jumped to conclusions that it was transmitting keystrokes, when "no no no, it merely has that potential, you see!" well, what do you expect?
- podperson 15y agoIt should be noted that (having finally watched the video) what he is accusing Carrier IQ of is key logging, but he does not demonstrate key logging of the keyboard (he demonstrates key logging of two of the hardware keys and the numeric phone dialer, but never demonstrates the same behavior when using the virtual keyboard -- he does show that the unencrypted url is being sent (which means that the Carrier IQ software has hooks into the browser where it shouldn't) but does not demonstrate key logging of, say, a password typed into a web page on an encrypted connection. If the article made this point it might have some credibility. As it is, it simply looks like Carrier IQ's PR department working overtime.
- drivebyacct2 15y agoIt's the same thing. The way virtual keypads work in Android and the same as the hardware keyboard. Also, it's already been noted that it is aware of HTTPS urls and reports those just as meerily (at least to the adb logcat, I can't make other claims). Further, the keyboard doesn't magically work differently when it's entering text into one page other the other... not sure what track you're thinking on there.
- podperson 15y agoYes, I agree that it's definitely tracking the url (and not the domain) but thet is not the same as key logging (which would capture things like passwords entered into a web page). I'll take your word for it that the text keypad is just as compromised as the numeric, but it would have been a lot more compelling to show a username / password (for example) being captured on a banking website before the user even pressed "submit" (or whatever).
- ghshephard 15y ago"There's zero evidence that Carrier IQ captured, recorded, or transmitted any keystrokes" ...except the 17 minute Youtube Video that showed Carrier IQ capturing, and recording keystrokes.
- spoondan 15y agoNo. Trevor Eckhart's video showed him capturing and recording the Android event log. This only demonstrates that the CarrierIQ agent running on the phone attached an event listener to keyboard events. While that behavior is consistent with a key logger, it's also consistent with lots of totally harmless software doing totally normal things.
- drivebyacct2 15y ago>it's also consistent with lots of totally harmless software doing totally normal things. lol. What "totally harmless software" or "normal things" require userland software to monitor every keystroke. And not just monitor them, but echo them so that they appear in the logcat?
- rictic 15y agoI don't think that they've earned the benefit of the doubt here, but I think it goes too far to say that there's no reason why benign diagnostic software might pay attention to key presses. For example, a particular button combination could bring up a diagnostics screen or send additional information. The logging, likewise could be just some debug code accidentally left in.
- marshray 15y agoHow about the https urls being intercepted? Those were for a diagnostic command sequence or debug code too?
- scott_s 15y agoTo be clear, they are monitoring the physical keys on the phone, not key presses on the virtual keyboard. The physical keys are the buttons for things like "Home" and "Back." Why you would monitor the physical key presses seems obvious to me: in the case of a phone or application crash, you want to know which keys were pressed at the time. In order to know that, you need to log it. It's possible that they are monitoring the key presses of the virtual keyboard, but I have seen no evidence of that. From what I can see, they're recording the same kind of information that, say, Chrome or Firefox record for crash reports.
- digitalboss 15y agoThis reminds me of the story that the media ran hard on for "...hackers launched a cyber attack against a water facility in central Illinois." More and more tech sites now are turning into these media engines that just want pageviews, their spinning a small story with very little facts into a Fox type news story, it's very sad to see, and will continue to happen, so they can be bought by the AOLs of the world. The Russian hack never happened http://www.tgdaily.com/security-features/59984-the-russian-hack-never-happened http://www.tgdaily.com/security-features/59984-the-russian-h...
- marshray 15y agoThe water pump story was primarily an elaborate screwup. This story has some very interesting reality under it and the details are still being uncovered.
- digitalboss 15y agoI agree with you, but with so little facts up front, and with seeing the story was so twisted in the media, it's just the reality of how things happen in today's world. Buckle up and hold on, innocent later.
- rmc 15y agoOne way to stop this is for us consumers of news to start getting more skeptical and rational. We should not gossip, we should look into stories before posting them to Facebook or upcoming them on HN.
- ChuckMcM 15y agoData really is an attractive nuisance. It is not too hard to see how many legitimate carrier questions could be answered to the benefit of consumers by having phones log data. Perhaps the simplest one is "Where do I have areas of weak coverage and so more dropped/distorted calls?" If a large percentage of the handsets can log GPS + signal strength info over a long period of time, the phones that are in suitcases (erroneously recording poor signal strength) and those in airplanes (erroneously recording great signal strength) will be balanced out by real world scenarios. Given 12 months of data (which covers various weather conditions, and correlating by date) one could quickly and efficiently decide where additiona transmitter/receiver resources should be deployed, and under what conditions it might make sense to bring up more resources and when you could shut them down (rain for example absorbs 1.5 - 2.5 Ghz radio signals. Of course that information can be accumulated much more painfully by driving around in a van to pick up this information. And the quality will be both better and worse, better in the sense that a signal-quality-van could have equipment to evaluate the signal strength very accurately, but worse in that you wouldn't get the time of day and conditional variation from weather. The 'attractive nuisance' aspect of it however comes into play when you've created a logging/monitoring platform and now you can collect other kinds of information. Like warrantless wiretappers who are unable to restrain themselves from listening in on conversations they have no reason to listen to, once the infrastructure is there the humans in the loop seem unable to prevent themselves from succumbing to their desire to 'know.' I had a long discussion with some law enforcement types at a security conference along these lines (I was advocating outright bans on some of the tools they consider 'essential' for fighting crime). The economic cost of not using the tools is higher (higher taxes (to pay for more cops), higher phone bills (to pay for signal strength vans)) and, until it is actualized, only a "potential" risk of abuse. That is such a terribly hard equation for people to evaluate on instinct.
- ozten 15y agoI'm calling B.S. on carriers don't charge for your key-logging data. My wire had some billing/technical issues and we worked through AT&T's data transmission logs, there were unexplainable uploads that seemed like telemetry or some kind of debugging. They denied it and told us it was our data uploads. Now I know what that probably was ;) Open up the tele-comm industry, this is really scare stuff!
- giberson 15y agoHow, in the same article, do you say "an expert examined the assembly of the software and determined there is no code that records or reports back" then follow that a little later with "carrier IQ doesn't make the decision, they provide the software with flags that the carriers can configure to record and/or report back".... These two statements don't really mesh up.
- marshray 15y agohttp://en.wikipedia.org/wiki/K%C3%BCbler-Ross_model http://en.wikipedia.org/wiki/K%C3%BCbler-Ross_model Stage 1. Denial
- dredmorbius 15y agoWho is Declan McCullagh and why should we believe he invented the Internet?