prepublication version of article Managing Intellectual Property, April 2005
Software is not an industry
Brian Kahin, University of Michigan
Accordingly, where an invention does not make a technical contribution to the state of the art, as would be the case, for example, where its specific contribution lacks a technical character, the invention will lack an inventive step and thus will not be patentable.
– Recital 12, Directive on the Patentability of Computer-Implemented Inventions
Hoping to get beyond the circularity of its discussions with software developers, the UKPTO has begun a series of workshops to “see if the definition evolving through numerous court precedents is the best one to formalise in the CII Directive.” Despite the apparently plain language of the European Patent Convention, the black-letter exclusion of computer programs “as such” has become a meaningless formality. As the European Commission explained in its 1999 Communication to the Parliament:
While computer programs are protected by patent in the United States and in Japan, in Europe we used a legal artifice: The programs per se are not patentable, while a technical invention which used a program is.
In the proposed directive, the Commission claims it would stop erosion towards U.S. practice by handing the line-drawing exercise back to the same experts that brought us problem in the first place. But as the Computer & Communications Industry Association recently wrote in a letter to the Council:
Experience in both the U.S. and Europe has shown that weakly defined limits invite relentless assault and circumvention. Weak limits are vulnerable to dominant providers, speculators, and intermediaries eager to expand the volume of their business.
Meanwhile the political struggle moves to the center stage of European politics in a standoff with the European Parliament on one side and the Commission and a divided and confused Council on the other. An extraordinary three-tiered request from the Parliament asking the Commission to restart the Directive is quickly denied without reason. National parliaments ride to the barricades on the side of the EP, but not quickly enough. Emotion runs high as the Council formalizes its decision with charges of manipulation and improper procedure – confronting the Parliament with a “compromise” version that is even further from the Parliament’s than the Commission’s original version.
The Commission’s refusal to rethink the Directive is remarkable in terms of how much its own thinking evolved from 1999, when it argued for following the U.S. lead, to 2002. Much has happened since 2002:
An explosion of software patents. In 1999 the Commission noted that Microsoft had accumulated 400 software patents. In 2004, Microsoft announced that it would file for 3000 patents in the current fiscal year alone!
The Federal Trade Commission held hearings throughout 2002 and issued a landmark report in October 2003 documenting the dysfunctionality of the patent system in the ICT sector.
Evidence of the massive liabilities that software patents can create and their allure not only for trolls but blue-chip companies such BT and Kodak. The 520 milllion dollar Eolas verdict demonstrates that even universities (who don’t need to worry about maximizing shareholder value) will use patents to claim basic Web functions if they can.
Recognition that software patents create liability for end-users and therefore threaten corporate and government acceptance of open source software. This has lead to competition based on indemnification, a game that belongs to a deep-pocketed few.
Recognition that the need for cross-licensing in complex technologies kills the exclusivity that patents give to small companies in discrete technologies.
The unveiling of Intellectual Ventures – heralding a new kind of cartel that cross-licenses its owners while it extracts what it can from the rest of the world. (http://www.msnbc.msn.com/id/6478691/site/newsweek/)
Not least of course is the Parliament’s strong reaction to the directive.
The Commission should not be bound by what it thought in 1999 or 2002, but by refusing to consider redoing the directive, it pretends that the world stands still. It alone among the three European institutions has the resources to monitor, analyze, and publish its insights. But it has done no more than stonewall the Parliament.
In today’s economy, patent practice and patent policy are clearly fast-moving targets. The debate is no longer just about drawing legal lines. It has taken on vast economic, political, and institutional dimensions. The Lisbon agenda is invoked on all sides. Dire predictions compete for scarce attention.
What is clear amidst the noise is that the patentability of software is a fundamental issue in innovation policy that once decided is unlikely to be revisited. The economic stakes are enormous. Software is not just an industry or sector, it is infrastructure for the world’s knowledge-based economy.