[cryptography] NSA's position in the dominance stakes

Steven Bellovin smb at cs.columbia.edu
Wed Nov 17 14:54:06 EST 2010

On Nov 17, 2010, at 1:23 12PM, James A. Donald wrote:

> On 2010-11-17 10:06 PM, Ian G wrote:
>> Yeah. And just on that question of patents, and so forth. IMHO, anyone
>> thinking that the patents aren't valid is ignoring the business risk of
>> the attack. An ethical or "unpatentability" defence is somewhere between
>> worthless and financial suicide :)
> They are only going after the people with big pockets, and the people with big pockets do not necessarily capitulate.

Capitulate?  The word is meaningless.  A big company goes through a very simple calculation: the patent owner (troll or otherwise) wants $X, tripled because of (alleged) willful infringement.  $3X is therefore at risk.  It will cost $Y to defend against the suit; if you lose, and the probability of that is Z, you're out $Y + $3X.  Even if you win, you're out $Y, and if you win on grounds of non-infringement (i.e., the jury felt that the patent was valid but you didn't use the protected technology), the plaintiff may be back 3 years from now to go after your next release.  Depending on Y, the uncertainty in Y, and the $X' << $X that the plaintiff will actually accept, you may be better off paying them to go away -- better off in a financial sense, since that's all that matters.  Corporations don't have moral principles, it seems; they have bottom lines.  You're not capitulating, you're making a business decision on how best to reduce your costs, even if they're just $Y.

As for small companies -- a win against a small company can be used later on.  The defendant may move for reexamination by the patent office, but once that's happened -- even if the challenge is poor -- the patent owner is in a stronger position; see, for example, the discussion of estoppel in https://www.eff.org/wp/patents-and-public-domain .  Beyond that, a win against a small company can be used in the economic damages section of a trial against a big defendant, as proof of the value of the patent.

		--Steve Bellovin, http://www.cs.columbia.edu/~smb

More information about the cryptography mailing list