This is why "responsible disclosure" is a joke. The flaws put in by these companies are not responsible. (Sometimes people make mistakes, but we're at the point of carelessness).
That said, the term "responsible disclosure" is Orwellian, and you should very much avoid using it.
The better term is "coordinated disclosure". But uncoordinated disclosure is not intrinsically irresponsible. For instance: if you know there's an exploit in the wild for something, perhaps go ahead and tweet the vulnerability without notice!
I see it as a kind of Hippocratic Oath in the field.
I must be misunderstanding. Would you mind expanding on this more?
I would personally in almost every case report vulnerabilities I discovered. But not in every case (for instance: I refused to report the last CryptoCat flaw I discovered, though I did publicly and repeatedly warn that I'd found something grave). More importantly: my own inclination to report doesn't bind on every other vulnerability researcher.
I understand morality is subjective, but that's my 2 cents on the matter.
EDIT: about the vulnerabilities you didn't disclose, I really can't understand why not. Why not just send an email to the maintainer: "hey, when I do X I cause a buffer overflow"? You don't even have to help them fix it. You probably won't answer this, but can you tell me why you wouldn't disclose a vulnerability?
I confess to being a bit mystified as to how work I do on my own time, uncompensated by anyone else, which work does not create new vulnerabilities but instead merely informs me as to their existence, somehow creates an obligation for me to act on behalf of the vendors who managed to create those vulnerabilities in the first place.
Perhaps you have not had the pleasure of trying to report a vulnerability, losing several hours just trying to find the correct place to send the vulnerability, being completely unable to find a channel with which to send the vulnerability without putting the plaintext for it on the Internet in email or some dopey web form, only to get a response from first-line tech support asking for a license or serial number so they can provide customer support.
Clearly, you have not had the experience of being threatened with lawsuits for reporting vulnerabilities --- not in software running on someone else's servers (which, absent a bug bounty, you do not in the US have a legal right to test) but on software you download and run and test on your own machine. I have had that experience.
No. Finding vulnerabilities does not obligate someone to report them. I can understand why you wish it did. But it does not.
Can't I just flip this around on you and say you have an ethical obligation to spend some of your time looking for vulnerabilities? If you started looking, you'd find some. Why do you get to free-ride on my work by refusing to scrutinize the stuff you run?
To some small extent, yes, though how much work is up for debate. Maintainer's email and PGP public key is right there on the website? Yeah, I think you're obligated. No email you can find, no way to contact them, or are just outright hostile? No, I think you shouldn't have to deal with that.
But I feel like you agree with that, though maybe not in those exact words. After all, you've had to jump through all kinds of hoops to disclose vulnerabilities, been threatened with lawsuits for doing the right thing, and yet you still practice responsible disclosure in almost every case in spite of the burden of effort and potential risk. Aren't you doing it because you think disclosure is the right think to do? That's all I mean by obligation.
EDIT: sorry, not "responsible disclosure," "cooperative disclosure" or whatever term you want to use for disclosing the vulnerability to the maintainer.
Nobody has to enter a burning car and risk his life but at least you have to call the emergency service or do whatever you can reasonably do to help. And it really doesn't matter whether you are doing your work delivering packages, whether the accident was the fault of the driver because he was driving intoxicated, if somebody else could also help or whatnot.
Discovering a vulnerability is of cause different in most respects - the danger is less imminent, the vendor may have a larger responsibility and so on. But the basic structure is the same - more or less by accident you end up in a situation where there is a danger and you are in the position to help to make the outcome probably better.
So I think one can not simply dismiss that there might be a moral obligation to disclose a vulnerability to the vendor on just the structure of the situation, one has to either argue that there is also no moral obligation in the accident scenario or argue that the details are sufficiently different that a different action - or no action in this specific case - is the morally correct or at least an morally acceptable action.
I would feel a moral obligation to help mitigate concrete physical harm to victims of an accident. I feel no such obligation to protect against hypothetical threats to computer systems.
Chances are, you recognize similar distinctions; for instance, I doubt you feel obligated to intervene in accidents that pose only minor personal property risks.
But I need to be careful saying things like that, because it is very easy for me to say that, because I don't spend any time looking for those kinds of flaws. Security research is pretty specialized now, and I don't do spare-time Windows work. I might feel differently if I did.
I would not judge the (many) researchers who would not necessarily disclose that flaw immediately.
So, it won't be fixed.
Hopefully only one or two people know about the same flaw you found...
Oh, but you would know ahead of time if concrete physical harm could possibly come to the victim of an accident?
Well good for you! You should probably be in charge of defending all infosec research, since apparently you can't be hacked.
A good shorthand definition for "responsible disclosure" is "report to the vendor, and only to the vendor, and disclose to nobody else until the vendor chooses to release a patch, and even then not until a window of time chosen by the vendor elapses."
Maybe you thought I was saying "the only way to disclose responsibly is to honor a formal duty to the vendors of insecure software". No, that was not my argument. If you thought it was, well, that's a pretty great demonstration of how the term is Orwellian, isn't it?
Or I could be missing part of your argument (it was quite terse, after all). Maybe you could fill in some details.
Thankfully, the black market doesn't want 99.99999% of the vulnerabilities people find.
I have friends who have sold vulnerabilities to people other than vendors. I do not think they're unethical people, and I don't know enough about those transactions to really judge them. So, it really depends, I guess. But if it were me, I'd be very careful.
I don't feel wrong saying that all of those are irresponsible. There are some people who write good code, who at least make an effort to avoid vulnerabilities, and those are the responsible ones.