Patents vs. Secrecy

giampy writes "New Scientist is reporting that the NSA appears to be having its patent applications increasingly blocked by the Pentagon. From the article: 'the fact that the Pentagon is classifying things that the NSA believes should be public is an indication of how much secrecy has crept into government over the past few years.'"

Re:win/win

[mfago]

Try to patent something that the government thinks should be (or is) secret and suddenly you'll find you no longer have any rights to it. Not sure if they are required to pay you, although Feynman eventually did manage to get $1 for the idea of a nuclear submarine...

Re:Compensation?

[xiphoris]

Or will they just say "Eminent Domain" and just take the thing

The principle of eminent domain does not allow the government to just "take" things. Eminent domain [wikipedia.org] requires that the government compensate you a fair market value [wikipedia.org].

Of course, that says nothing about other methods they have of preventing you from releasing your invention (national security?) or who decides what "fair market value" is.

Re:What The Post Doesn't Say

[Anonymous Coward]

Since November 29th, 2000, the USPTO has been required to publish all patents after 18 months of recieving them, or on acceptance, whichever is first. So yes, the PTO is being less secretive. I think the article is talking about patent applications that don't even make it TO the PTO, and are stopped before they leave the NSA/DoD.

Re:Secrecy

[geomon]

Having done a smidge of work for the government, I'm happier with secrets "just in case" than creating holes that might not have to have been made.

I'm sorry, but this attitude just smacks of laziness on the part of a classification clerk. When I worked at Department of Energy sites I was amused to discover that groundwater well construction documents known as 'as-builts' were classified during the Cold War. We had to send over a guy with a clearence to review the well log and report back to the classification clerk that no national security information would be disclosed by declassifying the record. At one site the DOE was custodian to over 4,000 wells, of which 90% of the records were classified. Every hour spent by a PhD geologist reviewing well records cost the government real money. This laziness in applying a classified status to well records cost the taxpayers millions of dollars throughout the DOE complex without advancing national security one iota. Countless other examples of construction records for other non-proliferation items were also classified.

Perhaps you like throwing money away for useless 'feel good' measures, but I don't.

Re:If you can't patent it...

[letxa2000]

the fact that the Pentagon is classifying things that the NSA believes should be public is an indication of how much secrecy has crept into government over the past few years.'"

What a load... "Over the past few years?" Come 'on, cut any silly implications that government secrecy is somehow something new with the Bush administration. The FOIA was passed for a reason and it was passed long before "a few years ago."

Government secrecy is nothing new... just the spin.

Re:Is DOD screwing up great NSA plans?

[LnxAddct]

The NSA, despite being secretive as hell, is one of the few government agencies that has consistently been upfront with the public. Multiple times they've found weaknesses in algorithms and fixed them, never giving an explanation, just a fix. In some cases it was years later that anyone started figuring out how exactly the changes worked to make the algorithm more secure, and some modifications still aren't understood by the public, but its been shown that they all increase the overall security of the algorithms in question. The NSA has motivation to make these as secure as possible simply because they also use these algorithms to securely exchange information among contractors and other agencies. I've read before that the NSA is as much as 50 to 100 years more advanced in mathematics than the rest of the world, now I don't know how accurate this is, but judging from their history it probably isn't too far off.
Regards,
Steve

Re:Compensation?

[ScrewMaster]

Not at all. This has nothing to do with eminent domain, this has to do with military secrets, and how the ability to peg something as "classified" results in the effective theft of intellectual property.

Back in the sixties, a company my father started did a lot of government contract electronics design and manufacturing, mostly for the Navy (some Air Force.) Some of his designs were parsecs beyond what the Navy was currently using at the time, so good that the Navy simply classified them outright. Okay, that's a compliment in a way, but it meant that he couldn't tell anyone about his concepts, couldn't use them for anything himself, and couldn't market any products made with them unless the government chose to buy them from him. Which they didn't, because after stealing his IP they simply shopped it around to other vendors to get a better deal (or to somebody's brother-in-law, whatever.) After that experience, he learned to withhold key parts of specifications so even if they classified what he gave them it wouldn't do them any good. He pissed off more than a few Navy engineers that way, but his attitude was simple: if it's good enough for the Navy to steal it's good enough for them to pay the inventor a fair price.

This all happened was forty years ago, and given the turn our society and our government has taken since, I can't believe the situation has improved any. Really, working for the military is a risky business for any private-sector operation, no matter how you slice it. Money to be made, sure, but you gotta be careful.

Re:If you can't patent it...

[Mattcelt]

cut any silly implications that government secrecy is somehow something new with the Bush administration

You're right, secrecy isn't a new idea in government. However, the sheer amount of secret things - classified data, blocked FOIA requests, and so much more has grown exponentially in the past 20 years or so. The amount of secrecy allowed in the US now is leaps and bounds above what it was when Reagan was president. (And it was a lot then!)

It used to be that data defaulted to "unclassified" unless it was specifically classified. But lately it's taken a quite a turn - more and more data is defaulting to "classified".

I think a large part of this has to do with two realizations at the government level. One, the less information about the government is out there, the less accountable their constituents can hold them. (This is why the FOIA is so critical for the protection of rights for US citizens.) Two, statistical mining, data interpolation and extrapolation, and other sophisticated, computationally-intensive information guessing techniques have advanced so rapidly and with such efficacy that even when only "non-sensitive" portions of data are released, people are becoming extremely good at figuring out the underlying secrets.

Personally, it scares me that the government can keep secrets from me without even telling me why they're keeping it a secret. ("National Security" has become the catch-all reason to classify ANYTHING, it seems.) It scares me more that the government will no longer let me keep secrets from it. That disparity is beginning to undermine the balance of power between the electors and the elect, and could very easily lead this country into a tryannical state. I thank God that there are still some idealists in the government who are trying to make the right decisions; it is they who help to counteract the creep of power and those it affects.

Re:I hope they give you compensation

[MstrFool]

No, they don't. They take it and run, use it how ever they like and hire any one they wish to make it for them. There is a case right now where that happened with some underwater cable connections. The guy is totaly SOL as being clasifide, he can't even show the evidence to a judge so it can't even go to court.

Re:There is not enough data...

[FellowConspirator]

FWIW, in 2004 the GAO cited a 95% increase in the number of government documents classfied as secret compared to the preceding 5 years. The same report also cites unprecedented levels of the executive branch refusing to provide requested documentation to congress and to withold information from the GAO itself including hundreds of instances of refusing to provide requested documents to the 9-11 commission including copies of presidential briefings on the subject.

The rate of FOIA challenges and denials has also skyrocketed.

Hmmm... Government getting more secret, AG writing memos about how torture is justifiable, enacting laws that permit indefinite incarceration without being charged, end to judicial oversight of wire-taps, congress considering a shield law for that would make it so only certain people can report government wrongdoing without threat of legal action... At this rate, how long will it be before the bill of rights falls into desuetude?

Re:If you can't patent it...

[Frnknstn]

Oops, that must have been why the filter kicked in :)

2001: 0 NSA Patents Blocked
2002: 0 NSA Patents Blocked
2003: 0 NSA Patents Blocked
2004: 5 NSA Patents Blocked
2005: 9 NSA Patents Blocked (up to March 2005)

RSA and GCHQ

[Flying pig]

I think it's well known that a GCHQ scientist (with the unfortunate name of Cocks) came up with the public/private key idea before Rivest,Shamir and Adelmann. British security predictably sat on it, with the result it was patented in the US and the UK lost the benefits of yet another bit of fundamental research.

However, given the prevailing attitudes in the English speaking world, I suggest you patent your ideas in the non-UK EU. Luxembourg?