Dear Google, Commit Your Broad Patents to a Defensive License
There has been quite a bit of news lately that you have applied for a patent on bill splitting. It's a helpful tool—we've all been to restaurants or lived with housemates and have needed a convenient way of calculating and tracking who owes what. But this technology already exists—and it has existed for a while. The fact that you feel compelled to file for this patent only further proves how broken the system is.
In fact, a number of companies have offered similar services over the last decade: Splitwise, Billpin, PayDivvy, WeSplit.it... the list goes on. The bill-splitting landscape not only drives the point that your patent should not be granted because of copious prior art, but it also highlights the many startups that would find themselves facing the fear of a potential patent lawsuit down the road. Because even if you claim you have no intent to assert this patent now, if granted, it will be hanging over the heads of businesses large and small for the next two decades.
On one hand, you, Google, are one of the companies leading the current push for important, meaningful patent reform to stop the onslaught of patent trolls and vague patents. On the other hand, you're asking the Patent Office to grant the very type of broad and vague patent you rightfully argue harms our innovation economy. (The co-founder and CEO of Splitwise recently wrote a blog post highlighting the disparity here.)
Now, we understand you may never assert this patent. Given your history, you may only be acquiring it "defensively." If this is true, we need you to practice what you preach, truly "increase freedom to operate and to reduce patent litigation," and commit to putting this patent under a defensive license. One viable option is the Innovator's Patent Agreement, created by Twitter, that asserts that a patent can only be used defensively unless the inventor gives express permission otherwise. You've also already put a handful of patents under a great non-assertion pledge aimed at protecting free and open source software from offensive patent suits; a stronger flavor of this would be welcome. Other inventive options like the Defensive Patent License, as you yourself have highlighted, are possibilities too.
If this patent application goes through, the grant would highlight two major broken elements of the patent system:
First, companies like yours are applying for bad patents, even if your intent is defensive. As Julian Sanchez put it so eloquently, "Good defensive patents are bad patents." They're pretty much definitionally broad (Your patent for machine learning on phones? Or Amazon's on digital first sale? Or Microsoft's on crowdsourcing!?), and nobody has any idea where they'll end up down the line.
Overbroad patents are like toxic waste—they need to be contained for the entire 20-year patent term to keep the public safe. A company may claim good intentions now, but a change in leadership or sale of the patent can quickly result in abuse. To take one recent example, American Express was granted a ridiculous patent on the security numbers on the back of credit cards. The company declared that it wanted this "technology" to be generally available and donated the patent to a non-profit. Yet, a few years later, the patent fell into the hands of the patent troll Intellectual Ventures and is now being asserted against AmEx's competitors.
Second, the Patent Office is granting these overly broad patents when an abundance of prior art clearly exists. They are already inundated with patent applications, and they granted over 68,000 software patents last year alone. This flood of software patents already spells out trouble for most startups: you're probably infringing something. But when broad concepts—like bill splitting—are patented, the landscape becomes even more hostile.
And we have a feeling you know that to be true. After all, we're pretty much on the same side of the patent reform debate. You yourself have decried "vague, overbroad" patents. So if this patent is granted—and if we weren't being clear enough, this patent should not be granted—we urge you to publicly commit it to defensive purposes only.
And while you're at it, do the same for your other patents as well.
Electronic Frontier Foundation