In the 1980s, Gary Reback was an attorney at Sun Microsystems, one of the hottest startups of its day. In a classic Forbes article, he described the day that a group of IBM patent attorneys visited Sun to demand that they license IBM's patents. They presented a list of seven patents that Sun allegedly infringed. Sun's lawyers inspected the patents and told the IBM lawyers that six of the seven patents were likely invalid. And Sun clearly hadn't infringed the seventh, they said.
"OK," an IBM lawyer responded, according to Reback. “Maybe you don’t infringe these seven patents. But we have 10,000 U.S. patents. Do you really want us to go back to Armonk [IBM's New York headquarters] and find seven patents you do infringe?" Sun wrote IBM a check.
This is the problem of patent thickets: when a large company holds so many patents that it becomes impossible to innovate without infringing numerous patents. Acquiring patents is a slow and expensive process, so incumbent technology firms will always have a lot more patents than up-and-coming firms. Patent thickets owned by IBM, Microsoft and other incumbent technology companies act as a tax on innovation, transferring wealth from today's innovators to the innovators of the past.
The problem has gotten worse since the courts lowered the bar on patent quality in the 1990s. Microsoft, for example, has been granted more than 20,000 patents in the past decade. That has enabled Microsoft to force 80 percent of Android vendors (by market share) to pay Microsoft royalties to use software created by Google. That's not because Google stole Microsoft's source code. Rather, it's because Microsoft has so many broad patents that every modern operating system infringes many of them.
Killing bad patents
A good first step toward cleaning up the mess would be to create an expedited process for invalidating patents that were granted erroneously. In 2011, the financial services industry convinced Congress to include an expedited process for invalidating business method patents in that year's patent reform legislation. But the provision was limited to patents affecting the financial services industry.
This year, Sen. Chuck Schumer (D-N.Y.) and Reps. Darrell Issa (R-Calif.) and Judy Chu (D-Calif.) have proposed expanding this program, known to insiders as the covered business method (CBM) program to cover many more patents. When defendants are accused of infringing low-quality business-method patents (which are often also software patents), they would have a new, less expensive process to get the patent office to invalidate them.
But companies that own a lot of these patents hate this idea. On Thursday, a coalition of companies and industry groups signed a letter to congressional leaders opposing expansion of the CBM program. The signers included Adobe, IBM, Microsoft, Qualcomm and Xerox. Adobe has received about 1,800 patents from the patent office, and the rest have thousands, if not tens of thousands, of patents.
"Subjecting data processing patents to the CBM program would create uncertainty and risk that discourage investment in any number of fields where we should be trying to spur continued innovation," these companies argue.
Yet many real-world innovators have the opposite perspective. Matt Levy is a patent attorney at the Computer and Communications Industry Association, which counts firms such as Google, Facebook, Samsung, Red Hat and Yahoo as members. Levy contends that the arguments of Microsoft and its allies are disingenuous.
"If your patents are valid, then you have nothing to worry about," Levy wrote in a Monday blog post. "The PTO won't review a patent under CBM review unless it decides that the patent is likely invalid."
"The signers of this letter aren’t worried about stifling innovation," Levy concluded. "They’re worried that their junky patents won’t be as easy to sell to trolls."
Congress tries again
On Monday, Rep. Bob Goodlatte (R-Va.), who chairs the House Judiciary Committee, released a discussion draft of a new patent reform bill. It includes a number of reforms designed to discourage patent trolling. It expands the CBM program, but only slightly. It doesn't allow a broad range of business method patents to be reviewed, as the Schumer and Issa/Chu proposals do.
The Goodlatte draft would make it harder for small, non-practicing entities to enforce their patents. Given how much these firms have abused the patent system, that's probably a good thing. But by itself, anti-troll legislation will merely cause the holders of broad, low-quality patents to look for different ways to profit from them. Trolling might become unprofitable, but it will be as profitable as ever for large companies to build packet thickets and impose an innovation tax on every new company in their industries.
To permanently fix the patent system, there needs to be an efficient process to invalidate the many patents that shouldn't have been granted in the first place. It's not a surprise that companies who own many of these patents oppose this idea. But as long as those patents are on the books, they'll be used against genuine innovators.