It's time for another fact check on the state of the Java-related patents Oracle is asserting against Google. Contrary to widespread misbelief, none of the six patents Oracle is still asserting (Oracle voluntarily dropped one of the original seven) has been invalidated on a truly final basis. By "trukly mean" I mean final-final, not just preliminarily final.
Depending on when the case goes to trial, Google could still be ordered to stop infringing any or all of those six patents. And even if the court stayed the case pending the completion of all reexaminations (when I say "completion", I mean the end of the whole process, not just the conclusion of the first part), statistics suggest that Oracle would likely have multiple patents to assert in the end -- probably not all six, but three or four are a reasonably probable number.
In two filings earlier this month, Oracle and Google updated the court on the status of the ongoing reexaminations. Three weeks ago, they summed up the state of affairs at that point, and last week, they reported that the USPTO had issued a "final Office action" (sort of "final", but not "final-final") concerning Oracle's '476 patent.
Google has done a great job on, and made considerably headway with, those reexaminations. It has clearly demonstrated that most of Oracle's asserted patent claims are of disputable validity. I wouldn't be surprised if Oracle at some point determined that it needs to assert some more of its 2,000 or so Java-related patents in a second (separate) complaint, and it's possible that the copyright part of this litigation has gained importance in Oracle's strategy at the expense of its embattled patent claims.
But again, not even one of those patents has been annulled so far.
Here's where things stand:
One patent (U.S. Patent No. 6,061,520) is considered patentable by the USPTO. This one will enjoy an enhanced presumption of validity at trial time.
There's one patent (U.S. Patent No. RE38,104, naming short-time Google employee James Gosling as the inventor) on which the USPTO hasn't commented yet, not even in the form of a first Office action. It's impossible to conclude anything from that fact. In its own report to the court, Google claimed that "[n]one" of the claims of this patent is "[c]urrently [a]llowed", but that's grossly misleading. There's a rather strong presumption of validity of issued patents under U.S. law (which in my view shouldn't be the case because the USPTO issues to many bad patents, but that's the law as the Supreme Court decided this summer). That presumption of validity is not even affected by first Office actions, but the mere fact that reeamination is ongoing, without anything known about the examiner's inclination, is lightyears away from a patent being less than (for the time being) valid.
For two other patents with respect to which there have been first Office actions (U.S. Patent No. 6,192,476 and U.S. Patent No. 7,426,720), the initial rejections have been confirmed by "final Office actions". However, if the USPTO says that it "closes prosecution", this just means that the examiner's work is done for the time being (it could be resumed later as a result of a decision to remand). It doesn't mean that the patent has actually been invalidated: Oracle can, and most likely will, appeal this decision to another department of the USPTO, the Board of Patent Appeals and Interferences (BPAI). I have seen different recent statistics about the outcome of BPAI appeals in such reexamination proceedings. Rejections by reexaminers are affirmed in a majority of cases, but they are reversed in their entirety in about a third of all cases and sometimes reversed in part, or remanded. Even the BPAI's decisions are appealable (to the Federal Circuit).
Let's assume, for the sake of the argument, that the Gosling patent also stands up pretty well. Then Oracle would have two valid patents and a good chance of reversing the decision, at least with respect to some of the asserted claims, with respect to some claims from one or two of the other four patents.
Oracle's argument for a near-term trial comes down to "justice delayed is justice denied". Google's argument is obviously that it shouldn't be held liable for infringement of patents of uncertain validity.
This is a high-profile litigation, and it certainly doesn't reflect well on the USPTO. It appears that it's rather easy to get a patent granted initially, but it also seems that the USPTO's central reexamination unit may make it too easy for those contesting the validity of an issued patent. One may wonder how much independent analysis is actually performed.
In a very recent filing, Oracle cited several cases in which courts determined that even "final" Office actions are prejudicial to a patent holder if shown to a jury in an infringement case while an appeal is ongoing. We will see what Judge Alsup decides. He can always tell Oracle that resource constraints force the court to stay the case unless Oracle narrows it to a very short list of patent claims, but as I explained on previous occasions, it would be very risky for Oracle to narrow its case too much.
I'm sure Oracle expected serious reexamination pressure when it started this litigation. You can't sue an organization the size of Google without significant blowback, and Google isn't just large and deep-pocketed but also very smart. Oracle didn't asserted seven patents because it thought it would prevail on all seven, but to maximize its chances of prevailing on some of them. At the same time it's possible that Oracle's decision-makers were much more confident of the strength of their patents when they started this than they are now when they look at the status of the ongoing reexaminations. Throughout the industry, it's possible that the value of these kinds of patents has been overrated. The difference between patents that have been issued but never tested and patents that have withstood massive challenges becomes clearer every day.
Given that so many issued patents are of questionable validity, plaintiffs will actually have to assert many more, not fewer, patents going forward. Companies that thought they sufficiently hedged their bets by asserting five or six patents may now find that even if they asserted ten patents, they might be lucky to succeed with only one of them. If they have many patents that read, at least potentially, on an accused technology, they might sue over 12 or 15 (or even more) patents right from the start and then narrow the case along the way, at a point in time when they know what invalidity contentions a defendant has presented.
If you'd like to be updated on the smartphone patent disputes and other intellectual property matters I cover, please subscribe to my RSS feed (in the right-hand column) and/or follow me on Twitter @FOSSpatents and Google+.
Share with other professionals via LinkedIn: