FBI Exploits Tor Browser Bundle to Target Child Pornographers
wired.com
wired.com
Did I parse the article correctly?
This is why I think TorBrowser should switch to Chromium. Whatever you have to say about Google, Chromium is unambiguously more secure than Firefox.
Chrome forces sandboxed processes to use its network stack; if that's been configured to use Tor, it shouldn't end up revealing your IP and therefore identity.
Mozilla acknowledges as much with the mission statement of their Electrolysis project:
https://wiki.mozilla.org/Electrolysis
Hopefully Firefox's situation will improve in the near future—at the very least, we'll all benefit if Chrome has a viable competitor. But in the meantime Firefox absolutely is behind Chrome in the security department.
(Aside from the process-per-tab model, what's Chromium got out of the box that Firefox still lacks?)
But the default browser bundle leaves JavaScript enabled by default for good reasons[1], and if you want higher security with JavaScript enabled and/or against non-JS exploits, Chrome's process sandbox is a big deal, and will probably keep it significantly more secure than Firefox for a good amount of time.
If you screw up and miss something, the FBI might not even need to exploit a bug in order to find users' IP addresses, they could just use documented browser features to get users to connect to their honeypot using their real IP addresses.
As a practical example, the FBI could slurp usernames and passwords from other sites the user has been to, then subpoena them to determine if they'd ever logged in without Tor, and so forth.
[1]: http://www.chromium.org/developers/design-documents/site-iso...
You could force all your traffic through Tor with a router. That way your compromised end point won't easily reveal your public IP. The attacker would be forced to generate some traffic they could fingerprint or use the microphone and camera to spy on you.
Who decides what is worth compromising basic security and a zero-day? In the recent past, the US federal gov't has shown itself far more hostile to those who challenged them, or ratted on their actions than actual dangerous people. Is the "similar to" being a whistle-blower? Maybe being person who criticizes the policies of the gov't while being employed by it? If I trusted my gov't to restrain itself to child pornographers and actual threats to the country (not government, country), I'd sleep easier.
Ultimately, lawmakers and the people who elected them. Additionally, elected presidents and the people who elect them, as well as the court system.
There's a whole lot of little details that make these easier or harder under certain circumstances, but none of these are impossible, especially if you have popular support for your cause.
If you don't like the way something works, you can talk to your congressmen, vote, support organizations that promote your cause, etc. Just complaining on the internet won't change much.
Obviously, I mean, it's fine if the government just walks into your house once a week for inspection because, pff, what have you got to hide?
And if all else fails, "think of the children" usually works.
The Freedom Hosting malware came later, and it's not clear whether there even was a warrant or not, since it all appears to be still secret. To argue whether the warrant was sufficiently particular is impossible to say. I can concede that was too large a scope, but that defense can't really be made until there's a defendant.
Was it even the same malware in both operations?
Let me name it: the name of that exploit is EGOTISTICALGIRAFFE.
Sound familiar?
They actually infected everything on that service provider, including Tormail. That was deliberate and part of an NSA op. They didn't hack the server, the hosting provider gave them root. Data seized is being secretly used in various Tormail operations (not covered by warrant), and in "parallel construction" in several drug-bust cases and one related to another leaker. I'm not saying this is a cover, but it is awfully convenient.
The targets are those who downloaded the porn, those who hosted the porn, but - as far as the article discussed - not the pornographers producing the content [1]
[1]Pornographer - one who produces pornography -- http://www.merriam-webster.com/dictionary/pornographer
TOR should stand on its own. If its security is dependent on legal framework than it's insecure.
In my opinion, nothing can ever be considered completely safe. All you can do is make sure the attacker's effort is bigger than the value of what you are protecting is in his view.
My God. Those pony-pirating monsters.
There are times when law-enforcement needs slightly more space to work than regular citizens; search warrants, for instance. The questions arise when we discuss where the line is for how much more space they need and when they need it.
99% of the time, I agree with you; law enforcement should have to operate under the same rules as everyone else. But, for that one percent, asking politely if any crimes are being committed on the premises is just not enough.
A given computer system allows access to a set of individuals who, according to some criteria, have been granted authorisation (by some party) to have access that system. This lists consists, at minimum, of:
1) The owner of the system
2) Other users who have been granted access by the owner of the system
3) Law enforcement agencies from one or more jurisdictions
If one were to implement this, then this would involve the software explicitly granting access to the law enforcement agencies in question. And if we assume this is desirable, then the next question to arise would be which agencies should be granted access.
Now, this is where politics and international law come in. Australia, for example, would love to have access to government systems of Indonesia. Similarly, China for the US. And arguably all permutations of these, and every country in the world. And in many cases they likely already do.
If you were to take the transitive closure of the desires and claims of all law enforcement agencies worldwide, then this would result in a situation where all such agencies have access to all system worldwide. If this is the case, then why even bother with exploits? It should be explicitly encoded in the default ACLs of all operating systems, perhaps using a regularly updated list obtained from a reliable source.
Microsoft, Apple, and the Linux kernel developers could then embed this list, and the necessary update mechanisms, directly into the OS. This would ensure that all agencies have access to the systems they desire.
The problem that comes here of course is that which agencies should be granted access to which systems. Should it be only those within the same country? Should it be those of allied countries as well? How do you determine whether country A should have access to systems in country B?
I understand there are legitimate needs of law enforcement. But I also, by default, take the position that computer systems should be secure, and that everyone who's work touches on this should be focused on ensuring this. This is the difference between white hats (the "good guys") and black hats (the "bad guys").
Disclaimer: I'm a computer scientist, not a politician, intelligence analyst, or international law expert. My views have also been influenced by the overreach of the NSA and it's partner agencies.
The internet is international. Governments aren't. What now?!
My own gut feeling is that we should try and head in the same direction as the email system. The internet largely conquered spam without needing government intervention or international treaties through automatic reputation systems, blacklisting of email abusers and so on. It's basically a giant group consensus system in which spam is defined as "whatever our users think spam is" instead of an explicit law. It works! Similar techniques could be applied to other systems like Tor, if only the developers cared (they don't).
Unless they have proper and signed warrant. Which is like any other citizen, eg. me or you, having a permit to go into some place X. If we have to follow those rules (e.g about unauthorised access) then they should too.
Breaking into computer systems (or houses) to find evidence is not playing by the same rules. Nor is getting that kind of power arbitrarily over everyone, as they do, without specifics about a target/case.
When we build software we have a responsibility to ensure that their use of this software - particularly for communication tools is secure.
If there are instances where a piece of software is not secure, I believe (and I gather the general opinion in the industry is) that we should do what is necessary to fix those problems.
But that isn't the FBI's job.
Well, if that's the analogy you want to use, then this is like police peeping in on the homes of thousands of people who have nothing to do with any crime, all in an effort to find a handful of criminals.
There are many people who are OK with that state of affairs.
I'm not.
If we accept that it's just fine for government agencies (and remember, it might not be a government in your own country) to find exploits, not report them in order to protect people who use that software, and keep those to themselves for their own benefit, then I have a problem with that. It leaves people at risk of being infiltrated not just by (some government agency) but by criminals e.g. looking for their credit card information.
Put another way, let's suppose you find a vulnerability in Chrome. Do you:
1) Report it to Google, wait until a patch is available, and then discuss it publicly (aka responsible disclosure)
2) Keep it for yourself, or sell to a government agency (possibly that of a country not friendly to your own) to take advantage of
The thing I really have a major beef with is that so many people think that option 2 is just fine. Software should be secure - everyone in the industry has a responsibility to protect users, not leave them vulnerable.
Though I hate to say it, a system that (accidentally) allows police forces to deanonymise select individuals whilst other actors cannot do so, and even the police cannot do so in bulk all the time, is not a terrible tradeoff. Snowden has made similar remarks, I believe. Of course the supply of zero-days is presumably not infinite. As pointed out, if Tor had more manpower they'd maybe switch to Chrome which is much harder to exploit. So this situation whilst it may not seem ideal is perhaps the frying pan, and as endpoint security improves we may find ourselves in the fire.
I recently proposed on tor-dev that individual relays be able to stop acting as introduction points for specific hidden services, with exactly this scenario in mind. They ignored me of course. They seem to feel invincible.
Can you give a brief explanation about why Chrome would be harder to exploit than Firefox is when used with Tor?