Charles Cella, CEO of BountyQuest, claims that by providing an incentive for Hunters to find this prior art, his company will help to strengthen the U.S. patent system. In fact, they have an ambitious mission: "Unleashing the Power of Knowledge; Improving the Patent System." Cella says this service has the potential not only to invalidate bad patents, but also to help verify good ones, presumably by finding prior art that is close but not close enough to invalidate."
 

How BountyQuest Works
The BountyQuest system is fairly straightforward. Bounty Posters submit descriptions of the information they need to invalidate patents, usually including the numbers of the patents they want to invalidate. Currently, this only involves U.S. patents, but Posters may post bounties on foreign patents as well. All active bounties are listed in detail on the site. 

Hunters register with BountyQuest. When they find art relevant to the subject of a bounty, they submit BountyQuest's online form to establish date and time of submission. They have 1 week to follow this up with the complete documents, preferably marked to point out their similarities to the posted patents. BountyQuest reviews the submissions and sends them on to the Poster. At the end of the posting period, the Poster rates submissions as to whether they have met the Poster's requirements for a "completely correct response." BountyQuest then audits the ratings and makes the final call on whether a submission is a match. If so, the first Hunter who sent in that submission (prior art) gets the bounty. 

Some basic rules: The prior art must be publicly available information — you can't submit your company's trade secrets, internal reports, or lab notebooks to win a bounty. Both Posters and Hunters retain anonymity, unless they give written permission to make their names available. Posters pay a $2,500 posting fee and also send their bounty money to BountyQuest, which holds the bounty money in escrow and pays it to the first Hunter who submits a single piece of prior art that matches all particulars of the patent in question. When a Hunter wins a bounty, the Poster pays an additional 40 percent of the bounty price to BountyQuest. 

BountyQuest is serious about the users' agreement that both Posters and Hunters must sign — it's a 12-page "legally binding document." This agreement spends a certain amount of verbiage saying that BountyQuest is a neutral forum for postings and submissions, it does not offer legal advice, nobody can hold BountyQuest responsible for the validity of any patents, the effectiveness of any prior art that a Hunter finds, etc., etc. The agreement also stipulates that in cases of disputes between Poster and Hunter, presumably as to whether the art the Hunter submitted qualifies for a bounty, BountyQuest will be the final arbiter. (Cella told me outside experts would be called on for close cases.) And the agreement has some interesting things to say about the use of prior art that Hunters submit, which I discuss in more detail later.

Professional searchers, used to in-depth interaction with their clients, may have some problems with this system. Because of the anonymity granted both Posters and Hunters, the Hunter must essentially abandon the reference interview and do the patent search the client originally requested, which is not necessarily the patent search the client really wants. There is in fact no way to determine what this client — the Poster — really wants. Also, the Poster may have a hidden agenda, another aspect that I touch on later. 
 

Payoff to Hunters
From the patent searcher's viewpoint, a Hunter on BountyQuest essentially does patent validity searches on a contingency basis — sort of a patent lotto. In order to be paid, you must find prior art that can invalidate the patent for lack of novelty, and you must be the first Hunter to find it. Also, you must satisfy the Poster's requirements through a single document. This has several aspects not in the Hunters' favor. For one thing, submitting prior art that strengthens a patent's position will not, for now, win a bounty, since by definition no one piece of art submitted will read on all aspects of the patent. Also, if several Hunters or even one Hunter submits multiple documents that,taken together, could invalidate a patent through obviousness, the Posters will pay no bounty. If the Poster has several pieces of prior art related to obviousness and wants just the last piece of the puzzle, then a single document that meets the need will win a bounty; but it must still be just one document. 

Cella told me this single-document requirement was intended to keep things simple at first. He welcomes feedback, and I suspect he will get some on this issue because of its bias in favor of the Posters. He did say he hopes to come up with a way "without unnecessary complexity" to reward Hunters who submit close but not exact matches.

Although hunting for BountyQuest might become a retirement job for power patent searchers, the rewards aren't all that great for a single successful hunt. Almost all the bounties currently on the site are worth only the minimum $10,000 (although BountyQuest itself has posted one for $50,000) — not a huge fee, considering that some of these patents support businesses worth hundreds of millions. And a Hunter will probably have to invest time and patent search expenses on quite a few of these postings before actually winning a bounty. When you consider that a normal validity search can cost thousands of dollars whether or not you find the killer reference, the Posters would seem to have the upper hand here. 

But a successful Hunter's reward doesn't necessarily stop at one bounty. I talked with Cella about one aspect of the users' agreement that has confused some potential Hunters. The agreement states that "the fact that a particular Submission matches the criteria of a Posting is a trade secret of BountyQuest. Hunters agree that they will not publish or otherwise reveal the nature of a successful Submission to any third party without the consent of BountyQuest." Cella emphasized that the "trade secret" does not constitute the art submitted for a successful hunt — just the fact that it won a bounty. 

On the one hand, there's nothing to stop BountyQuest from selling the same art to other interested parties. In fact, they state clearly in their users' agreement, "In some cases, BountyQuest may have the opportunity to resell the information that a Hunter submits, such as to a third party.... You agree that BountyQuest has the right to resell any information you provide to it." In other words, BountyQuest may give the art to the Poster who paid the bounty, and then turn around and sell that art to a third party for millions. But there's also nothing to stop Hunters from giving the same art to their own patent attorneys — or selling it to other interested parties — as long as they don't identify it as having won a bounty. 

And in fact, if you do win a bounty, you could win more bounties with the same prior art. When a bounty hunt has a successful conclusion (bounty paid), BountyQuest plans to post that fact on their Web site, keeping Poster and Hunter anonymous, of course. Any other company that wants to invalidate the same patent can post another bounty on it, and the winning Hunter is free to send in the same art to win another bounty. And another, and another....

Why should multiple companies post bounties on the same patent? Consider this scenario: Company A has a patent covering a very profitable widget. Company B considers Company A's patent invalid for lack of novelty. Company B of course wants to find the killer prior art, a patent or publication describing that widget and published before Company A filed its widget patent. However, Company B does not necessarily want to try to invalidate Company A's widget patent — court cases can be so long-drawn-out and expensive and uncertain in their outcome. When Company B finds the appropriate prior art, it is much more likely to start making the same widget and wait for Company A to send accusations of infringement. Company B then quietly slaps Company A with the prior art, and Company A may well back off rather than risk having its patent invalidated.

You should also realize that Company B may not be anxious to share its prior art discovery with all the other widget manufacturers. After all, if only Company B knows about the prior art, Companies A and B will likely be the only ones to make that particular widget. But if Companies C, D, E, and so on find out about the killer prior art, then suddenly lots of competitors could start making the widget.

So, bottom line, when a Hunter wins a bounty for finding prior art against the widget patent posted on BountyQuest, all the other widget manufacturers may sit up and take notice. They will realize that prior art does indeed exist that the initial Poster considered killer prior art. Cella predicts that these other companies will promptly post their own bounties to learn about the same prior art. If 20 companies are interested in killing the same patent, all 20 of them might post bounties once they think someone has found the prior art they need. The same Hunter could win all 20 bounties; or each of the 20 bounty hunts could become a race between several Hunters to submit the same prior art first.

One more point, however: The original Poster has the right to forbid BountyQuest to announce that someone won the bounty. And the original Poster, in Company B's position, may well decide to forbid it. I wonder how many successful bounties will in fact be announced on the site?
 

Bounties on the Web Site
I asked Cella whether he thought the bulk of the Posters would be large companies or smaller entities. He thinks they will come from all over the map and will even include individuals passionate enough about an issue to be willing to post a bounty on it. (And pay $14,000 minimum!) 

So far, BountyQuest offers a limited number of bounties. Most of them fall into two major categories: biotech/pharmaceuticals (including the patents allegedly for Viagra and minoxidil) and software/Internet/business methods (including Amazon.com's "One-Click" patent and Priceline.com's reverse auction patent). One of the investors in this new company is Jeff Bezos, founder and CEO of Amazon.com, and one of the inventors listed on the One-Click patent. Another is Tim O'Reilly, head of O'Reilly & Associates, publisher of books on software and the Internet, and a vocal critic of Internet patents. O'Reilly in fact posted the bounty for Amazon's One-Click patent. BountyQuest has posted two bounties on its own, one of them for its own way-of-doing-business technology (patent not yet issued) — a challenge for the skeptics, presumably.

An interesting question arises vis-à-vis some of these postings. My good friend Edlyn Simmons, who for many years supervised the patent information function at Hoechst Marion Roussel, tells me that the descriptions on some of the pharmaceutical patents posted don't match the actual patents. As I mentioned, two of the pharmaceutical bounties are for patents described as covering minoxidil and Viagra (although the latter posting no longer refers to that product by name). Edlyn says, "These patents don't cover minoxidil and Viagra. The 'patent' on minoxidil expired ages ago.... The patent they're trying to invalidate is a combination of minoxidil with an antisense oligonucleotide, a very narrow invention that isn't approved by the US FDA (it's not in the Orange Book). The patent on Viagra, which I just looked up in the FDA Orange Book, is US 5250534, which you can compare with the patent in the BountyQuest site [US 5981563]. That patent talks about phentolamine, not Viagra (sildenafil)."

Do some of the Posters genuinely not know what the patents they are posting really cover? Or are they posting patents they truly want to invalidate, but adding misleading descriptions for reasons of their own? Are the Posters aware that some of the patents posted are just the tip of the iceberg — that the companies holding these patents have often filed whole hedges of patents covering the technology in all its aspects? 

Unquestionably, the Hunter has an advantage if he/she knows enough about both the technology itself and the patent art in the technology to be able to read the posted patents knowledgeably and make the sorts of calls that Edlyn did. However, Cella assured me that, whatever their actual coverage, the patents posted are the ones the Hunters should aim for. If a Hunter submits art that reads directly on the patent posted, then BountyQuest will pay the bounty, even if it turns out that invalidating that particular patent does the Poster no good. 
 

Who Are the Hunters?
I found it interesting that despite a flurry of newspaper articles published within a day or two of BountyQuest's launch and some trade press pieces within a few weeks, nobody from BountyQuest posted anything on the Patent Information Users Group (PIUG) discussion list. (Information veteran Bob Buntrock, who now lives in Minnesota, first saw a story in the Minneapolis Star Tribune and notified the rest of us via the PIUG list.) Cella assured me that he does indeed want to reach professional patent searchers. To this end he had talked with a number of independent patent searchers in the DC area who search at the USPTO, and he also made an announcement at an AIPLA (American Intellectual Property Law Association) meeting. What's interesting is that nobody he talked to thought to refer him to PIUG, or indeed had ever heard of PIUG. (Apparently we in PIUG have some outreach to do!)

However, Cella thinks good submissions may come from outside the professional patent searchers' arena. He believes that scientists and engineers working in the field can find what is needed from their own knowledge of what has been published. The Web site gets quite enthusiastic on this point: "Do you do cutting-edge research, or do you know people who do? Are you interested in the latest developments in technology or the life sciences? Maybe you read about the information we're looking for in a trade journal 10 years ago — or wrote about it in your doctoral thesis last month. Look for bounties in your field of research; harness the power of your knowledge, and get paid for what you already know. Do a little footwork, Web work, library work — whatever it takes to find the reference you know is out there.... The collective minds of tens of thousands of researchers can yield more useful information in one day, than hand searching at the USPTO [United States Patent and Trademark Office] can deliver in a month." 

In other words, the BountyQuest people hope to reach such a large cadre of Hunters that sheer luck will let them find the killing references that have eluded the professional searchers who tried to invalidate the same patents before they were posted for bounties. This is certainly possible. The professional searchers will find the art if it's in the commercial databases. But so much that constitutes prior art never appears in databases, and someone else might be lucky enough to find it. If a single copy of a master's thesis resides in a university library in Argentina, that's prior art — if someone remembers it and looks it up.
 

Patent Lore on the BountyQuest Web Site
The people at BountyQuest believe that Hunters need not have much familiarity with patents before using the site. Professional searchers, many of whom are registered patent agents and most of whom know a lot about patent law and practices, might take issue with that. However, the site provides a great deal of information to help educate new Hunters. In fact, it has quite a useful compilation of patent lore. 

For starts, BountyQuest offers a tutorial on the concepts of patent validity, what constitutes prior art, and where to look for it. This section talks about the difficulties of reading patents, especially the claims (which it calls "checklists"). It describes its postings as "Cliff's notes for patents" and says they provide the required reading without the legal jargon. (Well, maybe —­ see above.) It also discusses how to look for prior art. ("Finding prior art is not too hard if you know where to look.") (?!)

Admittedly the tutorial is a bit simplistic. But the Patent Information Center is well written and provides all sorts of useful patent lore. The Q&A section covers what a patent is, what can and can't be patented, the lifetime of a U.S. patent, the concepts of validity and infringement, and a few differences between U.S. and other countries' patent systems. The Patent Basics section goes into more detail on the three requirements for a U.S. patent: novelty, nonobviousness, and utility. It then talks at some length about Internet/software/business method patents and biotech/genomics patents. 

One long section discusses the whole concept of intellectual property — patents, trademarks, copyright, and trade secrets. This site includes links to WIPO (the World Intellectual Property Organization) and the Library of Congress Copyright Office.

There's also a fascinating history of U.S. patents that discusses, among other things, why patents "fell out of fashion" in the 1930s. FDR apparently had little use for the patent system, seeing it as a cause of the "economic malaise gripping the country." And the courts tended not to recognize patents as valid. But that changed in the 1970s, as international competition increased and U.S. companies saw that Europe and Japan offered stronger patent protection than the U.S. When the newly formed Court of Appeals for the Federal Circuit (CAFC) started hearing patent cases in 1982, patents gained new strength. Then came the advent of biotech patents, software patents, and Internet business method patents. As I say, it's interesting reading.

Finally, there's a glossary of patent-related terms and phrases. Want to know about Doctrine of Equivalence, injunction, prosecution history estoppel? It's all here.
 

Conclusion
Patents are in the news these days, especially those related to the Internet and ways of doing business. The Amazon.com Web site includes an impassioned letter from Jeff Bezos outlining reforms he would like to see in the U.S. patent system vis-à-vis business method and software patents. He feels these patents should have a lifetime of only 3-5 years, and that this should be retroactive — currently active business and software patents should come under this law. He also feels that such patents should have a 1-month public comment period before being issued to let the Internet community and interested others submit additional prior art to the patent examiners. Cella admits that the chances of such a law actually passing are fairly slim. So he hopes that BountyQuest will at least help to strengthen the patent system by increasing the likelihood that someone — preferably (in his view) BountyQuest Hunters — will find the prior art to cull the weak patents. 

Certainly being a Hunter can have benefits above and beyond the bounties per se. Successful Hunters may announce that they have won a bounty, as long as they don't reveal the art submitted, which would certainly provide good publicity for independent patent searchers looking for clients. And, as Roy Zimmerman points out, the concept of going after a bounty gives one a certain feeling of cockiness, of strapping on one's guns — "It lets us walk with a new swagger, eh?" 

It remains to be seen whether this site will succeed. Since postings last a minimum of 2 months, and bounties are not paid until the end of the posting period, BountyQuest won't pay any bounties until late December at the earliest. But if all goes well, BountyQuest plans to expand beyond patent-infringement information to other hard-to-find information, using the power of the Internet to link the people who need it with the people who can find it.