[tptacek]

Contrary to popular opinion (because "first to file" sounds insidious), whatever marginal effect this change is going to have will likely be positive. The intent of the change is to harmonize the US with EU on a corner case of patent law, where two entities file for a patent on the same invention during the same window of time. The new rule says that instead of clubbing each other over the head with lawyers, the first filing simply wins. More importantly, the rule changes strengthen prior art challenges to applications.

[lambda]

I think the issue that's concerning is that this increases the necessity to obtain a patent on any novel idea, lest someone else copy your idea, patent it, and then sue you for implementing your own idea. This means that now, every time you come up with some new idea, you need to spend all the time, money, and effort to patent it (and hope, in the process, that you don't come across an existing patent for that idea, since now your infringement of that patent will be willful with treble damages).

Especially in the software field, this causes all kinds of problems. As many people have pointed out, if you were to try and patent software that you create, you would have to file a patent for every few lines of code you write. With first-to-invent, I would expect (though I'm not a lawyer, and haven't studied this in depth, so I could very well be wrong) that you would be able to simply not bother patenting it, and then if sued, file a patent and use the first to invent rule to get your patent to override theirs.

[tptacek]

It does not mean that. The filer has to prove they invented the idea, which they can't do if you've published it prior to their filing.

If you weren't going to publish or file a patent, nothing changes for you at all; you're exactly as exposed to patent litigation as you were prior to the change, because your inaction was no more effective at blunting bad patents under "first to invent". If you invent something and keep it a secret, other people have always been able to patent the same idea, because the law does not require people to read your mind.

Again, I think the issue is that the "move" from "inventing" to "filing" conveys a kind of paperwork urgency that just isn't there.

[DannyBee]

"If you invent something and keep it a secret, other people have always been able to patent the same idea, because the law does not require people to read your mind."

This is not right. Prior to AIA it was possible to invalidate a patent by showing secret invention by another inventor prior to the patent invention date, among other things.

See also: Metallizing Eng. v. Kenyon (secret commercial use by party) or Eggbert v. Lippman (secret use in public) or Gore v. Garlock (Secret use by third party).

In fact, there is a bunch of "secret prior art" that AIA eliminates.

[tptacek]

Whoah. How did that work?

Also: from having been through the patent process a bunch of times: does this make it "harder" to obtain a patent by eliminating the ability to sell the invention under NDA before filing? I remember this being part of the M.O. of building, selling, and then patenting technology at companies I've worked at.

[DannyBee]

There were a variety of forms of secret art (in the sense that the inventor could not possibly have known about them when they filed for the patent), they were only brought up in re-exams or trials, of course.

There are also some sad cases where out-of-order patent processing caused later filed patents issued more than a year before earlier filed pending app. You don't get interference proceedings if the later-filed patent issued more than a year before your pending app. You just get a huge mess instead ;)

To answer the other question, since AIA makes secret commercialization not prior art, I expect more people will sell stuff under NDA[1]. Hopefully, more people will publish too.

[1] The statute itself is a bit ambiguous, but the PTO's new guidelines say "The legislative history of the AIA indicates that the inclusion of this clause in AIA 35 U.S.C. 102(a)(1) should be viewed as indicating that AIA 35 U.S.C. 102(a)(1) does not cover non-public uses or nonpublic offers for sale "

[riemannzeta]

Your analysis seems not quite right. Metallizing, for example, is a public use case in the sense that the legislative history uses the term even though the public could not have understood how to make the invention from the public use. I think the legislative history here is meant to say that the clause is meant to have the same scope as the old 102(b) bars except for territoriality. There's stil the new clause though...

[michaelfeathers]

The unstated assumption here is that the patent office actually does the work to see whether anything related was published prior to granting the patent.

[kleiba]

It does not mean that. The filer has to prove they invented the idea, which they can't do if you've published it prior to their filing.

But wouldn't that be exactly "first to invent" rather than "first to file"?

[nullc]

It's patents, so the terminology has to be confusing!

The "First to" is speaking about what happens in an interference proceeding between two patent applicants. What counts as prior art is largely unchanged— it's arguably even strengthened a bit (as a dishonest player could previously get away with more backdating of his lab notebooks).

Third party— that is, not derived from the applicant's work— public practice (which is pretty broad, including publication and use in a product available to the public) is an absolute bar to patentablity. After one year the "not derived" part goes away and even the applicants own publication (and work derived from it) constitute prior art. (this was all also the case before, though the details a somewhat subtly different)

To quote the statute:

(a) NOVELTY; PRIOR ART.—A person shall be entitled to a patent unless—

(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention; or

[...]

(b) EXCEPTIONS.—

(1) DISCLOSURES MADE 1 YEAR OR LESS BEFORE THE EFFECTIVE FILING DATE OF THE CLAIMED INVENTION.—A disclosure made 1 year or less before the effective filing date of a claimed invention shall not be prior art to the claimed invention under subsection (a)(1) if—

(A) the disclosure was made by the inventor or joint inventor or by another who obtained the subject matter disclosed directly or indirectly from the inventor or a joint inventor; or

(B) the subject matter disclosed had, before such disclosure, been publicly disclosed by the inventor or a joint inventor or another who obtained the subject matter disclosed directly or indirectly from the inventor or a joint inventor.

[sp332]

Multiple people can independently invent the same thing, either over time or all at the same time. Whichever one files first, wins.

[pasbesoin]

One of my concerns is that it gives the upper hand to large organization with the resources (including perhaps labor costs fixed to a certain degree by virtue of staff attorneys; also, filing fees) to file early and often.

Further entrenchment of IP as the purview of those who can afford it.

I'm glad to be proven wrong -- I'm not well-informed on this. But I fear it, both from my own limited thought exercise, and out of concern for the types of legal changes that are able to make it through the Federal process, these days, generally speaking.

[cube13]

First-to-invent also gave the upper hand to large organizations that had the manpower/resources to document every step of their invention process.

In all honesty, it's more of a wash.

[shmerl]

Parallel invention by different parties should simply be a sign of unpatentability to begin with.

[dj-wonk]

I'm chucking because you just used "simply" and "unpatentability" in the same sentence. Maybe you can argue your point in a probabilistic sense, but I don't see how you can do so based on the underlying principles. Parallel inventions can certainly still be novel, useful, and non-obvious.

What if we think bigger and better? I can envision a more perfect patent system protecting two (or N) parties who independently invented the same thing. After all, a key justification for patents is to make the up-front R&D cost worthwhile. In the case of parallel invention, it seems more fair to the parties and the public interest to not force N valid inventions to result in either zero or one patents.

[shmerl]

It should be a matter of definition. Since software patents don't deserve to exist altogether, making two parallel invention unpatentable would be only weaker, but still some positive development. I'd argue that this logic should apply across all patent field, not just for software, but for software the benefits are the most apparent.

[astrodust]

Although it triggers a frothing at the mouth sort of anger that it's now a race to file first, a more careful reading of it suggests that your patent should be denied if any evidence of your invention occurs before you file for it.

That is, one-click shopping cart type patents would be automatically invalid if any such thing already existed or this sort of mechanism was published or discussed by another party. Is this correct?

[tptacek]

Hasn't that always been the case? That's why when people start asserting bad patents, the Internet lights up with quests to find prior art to invalidate the patent.

[astrodust]

Usually this is after the patent has been issued which is very difficult to unravel.

Plus even if there is prior art, the entity applying for the patent could assert their "invention" came prior to the prior art.

[jordanb]

I think another important point is that it removes a lot of potentially expensive litigation and discovery on who was actually the "first to invent" from a patent case.