Select your localized edition:

Close ×

More Ways to Connect

Discover one of our 28 local entrepreneurial communities »

Be the first to know as we launch in new countries and markets around the globe.

Interested in bringing MIT Technology Review to your local market?

MIT Technology ReviewMIT Technology Review - logo


Unsupported browser: Your browser does not meet modern web standards. See how it scores »

{ action.text }

Clouds Over Crystal City

Just because software hasn’t experienced a cyber-Bhopal doesn’t mean it won’t ever happen. Indeed, the noxious clouds of litigation now gathering around e-commerce are renewing industry fears. What’s more, in what lawyers are calling the “post-State Street environment,” all bets are off for a software dŽtente as companies are not just patenting specific algorithms, but far more valuable Internet business concepts. And to the great dismay of critics who fear a wave of crippling lawsuits, the PTO may be granting many of these software patents in error, simply because it can’t keep current with advances in the field.

By law, no invention can be patented that has already been patented by someone else or has been published prior to the time the patent is filed; in the language of the legal system, such patents and publications are known as “prior art.” A key problem is that software programming-especially in its early days-was famous for its lack of a published paper trail and for the informal exchange of code and techniques among programmers. These poor “non-patent” records, combined with the PTO’s late arrival to the software game, mean the agency examiners who scrutinize applications often have tremendous difficulty establishing exactly when an invention was first made.

“The prior art problem is one issue where software patent proponents and opponents typically find common ground,” says Alan Fisch, who in 1994 was one of the first computer scientists hired by the PTO to work on software patents. Although the PTO has hired hundreds of new software examiners since then, Fisch says that, despite the agency’s best efforts, “the corpus of existing software patents does not define the totality of software innovation.” In other words, the PTO’s collection of software art still resembles the tip of the computer science knowledge iceberg. Although the PTO boasts that examiners have access to some 900 online databases, the reality is that, by almost all accounts, prior art searches in the software field sorely test the PTO’s capabilities.

The best-known example of the PTO’s mishandling of prior art came in 1993 when the California-based firm Compton’s New Media, creator of an early multimedia CD-ROM titled Compton’s Interactive Encyclopedia, announced that it had received a patent on multimedia itself-specifically, the “process and concept” of retrieval technology in multimedia databases. With scores of multimedia CD-ROMs already on store shelves, the announcement was a bombshell.

The patent seemed ridiculous, not least because techniques for indexing and searching multimedia databases had been explored at Xerox’s Palo Alto Research Center almost two decades earlier. Under pressure from the software industry, Bruce Lehman, then-commissioner of the PTO (Dickinson’s immediate predecessor), took the highly unusual step of second-guessing his own examiners. He called on the Patent Office to re-examine the patent, this time taking into account “new evidence” that had come to light-namely, the prior art the agency had missed the first time around. The result: Every one of Compton’s 41 claims were rejected.

Over the past decade many have tried to redress the prior-art problem. One group of programmers, led by Bernard Galler, now professor emeritus of electrical engineering at the University of Michigan, founded a venture in 1992 called the Software Patent Institute. The idea was to have programmers voluntarily submit a pool of information about existing know-how that the U.S. Patent Office could use in its search for prior art. While the PTO’s ability to detect spurious patent claims has improved through such undertakings, Galler admits the effort has met with only limited success and that there is still far to go.

Just how far remains an open question. One of the Patent Office’s harshest critics, Greg Aharonian, publishes the Internet Patent News Service and makes his living investigating the validity of software patents on behalf of firms involved in lawsuits. Aharonian claims that half to 70 percent of the software patents issued can’t pass what he calls “the Crystal City test.” That is to say, if U.S. patent examiners ventured beyond the walls of their headquarters in Crystal City, Va., they would find that the techniques they are patenting are already widely known and used among programmers.

“The Patent Office is doing a horrendous job in the software area,” Aharonian contends bluntly, “and they have made zero progress since the Compton’s New Media patent in handling non-patent prior art.” In a 1999 survey, Aharonian found that half of all patent applications surveyed cited no prior art at all. The result, he believes, is a disaster in the making. “No matter how bogus they may seem, patents can be powerful weapons,” Aharonian says. “As people start making money on the Internet, you can fully expect these patents to start being asserted.”

0 comments about this story. Start the discussion »

Tagged: Business

Reprints and Permissions | Send feedback to the editor

From the Archives


Introducing MIT Technology Review Insider.

Already a Magazine subscriber?

You're automatically an Insider. It's easy to activate or upgrade your account.

Activate Your Account

Become an Insider

It's the new way to subscribe. Get even more of the tech news, research, and discoveries you crave.

Sign Up

Learn More

Find out why MIT Technology Review Insider is for you and explore your options.

Show Me