Vortrag von Gregory Aharonian
1 Fr¨here Artikel 1
2 Ger¨st des Berliner Vortrags 2
2.1 Wem n¨tzen die SWPatente? Was f¨rdern sie? . . . . . . . . . . . . . . . 2
2.2 Wem schaden die SWPatente? Was hemmen sie? . . . . . . . . . . . . . . 2
2.2.1 Neuland betretende Firmen . . . . . . . . . . . . . . . . . . . . . . 2
2.2.2 Aktion¨re . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.2.3 Opfer von Schutzgelderpressungen . . . . . . . . . . . . . . . . . . 3
2.2.4 Inhaber untersch¨tzter Patente . . . . . . . . . . . . . . . . . . . . 3
2.2.5 Flinke Unternehmer, deren Arbeit verz¨gert wird . . . . . . . . . . 3
2.2.6 Neugr¨ndungen in patentvermintem Gel¨nde . . . . . . . . . . . . 4
2.3 Sind die Patent¨mter ihren Aufgaben gewachsen? . . . . . . . . . . . . . . 4
2.4 Verpassen die Europ¨er Chancen? . . . . . . . . . . . . . . . . . . . . . . 5
1 Fr¨here Artikel
Aharonian’s Statistiken werden gerne von Softwarepatent-Apologeten zitiert. Peter Han-
na, Software-Referent der UNION1 , begr¨ndete damit auf dem M¨nchenr UNIONs-
Konferenz 1997 seine zentrale Botschaft “The Debate is Over – Industry Demands Soft-
ware Patents” und empfahls sie u.a. mit den Worten:
The present International Patent Classiﬁcation System does not allow for
any easy analysis of the most common subject areas for Software Patents, or
to ﬁnd some of the new emerging software technologies. The best analysis of
this that I have seen has been done by Gregory Aharonian.
• 1998 Software Patent Statistics: Who is getting the Patents?2
• Inadequacy of Prior Art Searches in US, EU and Japan3
• trivial/obnoxious business method patents4
• The Patent Examinations System is Intellectually Corrupt5
2 Ger¨st des Berliner Vortrags
2.1 Wem n¨tzen die SWPatente? Was f¨rdern sie?
Software patents basically help two diﬀerent types of companies. It helps small companies
that actually have invented something new in the world of software (a hard thing to do
given the tens of millions of programmers around the world) and need protection from
large competitors as they enter the market with their product. One of the few examples
of this was when Stac won $90 million from Microsoft after Microsoft infringed Stac’s
disk ﬁle compression patent. But large settlements like this are rare (and some argue
that Microsoft should have won the lawsuit).
The other group of entities to beneﬁt are very large companies like IBM, which with
their portfolios of hundreds and thousands of software patents, can obtain hundreds of
millions of dollars, if not billions, from forcing smaller and newer companies to license
part or all of their portfolios. This is a intergenerational transfer of wealth.
2.2 Wem schaden die SWPatente? Was hemmen sie?
2.2.1 Neuland betretende Firmen
First, the smaller and newer (or new entrants such as European companies) compa-
nies having to pay royalties to older companies with well established patenting systems
and large portfolios of issued patents. These royalties are monies not then available to
compete in the marketplace, sometimes against these larger and older companies.
A second set of losers is stock market investors. Increasingly companies getting software
patents are touting them in the press and on the Internet, even if the patent is of
questionable validity. This leads a naive investing public to bid up the price of the
company’s stock. The stock price eventually drops as the market determines that the
company’s patent is probably of little value as a strategic weapon. Those investing too
late thus lose money as the stock price readjusts to a more realistic valuation.
2.2.3 Opfer von Schutzgelderpressungen
A third sets of losers is large numbers of small and large companies being forced to pay
monies to avoid patent lawsuits. For example, if someone with a patent of questionable
validity oﬀers you a universal license for $30,000, it is hard for a company not to pay
the $30,000 because to do anything else (starting out with obtaining a invalidity opinion
from their lawyers) will cost as much if not more. The $30,000 becomes a cheap “insuran-
ce policy” (compared to the alternatives), even though the patent may be completely
invalid. Some argue that companies asserting such patents are engaging in little more
With the US Patent Oﬃce issuing over 20,000 software patents a year, companies
could easily soon be in a position to have to pay these fees a few times each year, say
$100,000 worth of licensing fees, money better spent on R&D and marketing. These
possibilities have lead to a growing market for patent infringement insurance.
2.2.4 Inhaber untersch¨tzter Patente
A four set of losers is those small companies with new software patents that are actually
valid, but presumed to be invalid like so many other software patents actually are in-
valid. These small companies thus have to spend more time and money trying to have
their patent(s) considered more valid than public perceptions assume them to be. Large
numbers of low quality patents diminish the importance and value of the small numbers
of high quality patents.
2.2.5 Flinke Unternehmer, deren Arbeit verz¨gert wird
A ﬁfth set of losers are companies that experience delays in entering a software market
while they assess the quality of patents that aﬀect the market. For example, Priceline.com
has a patent on its reverse auction technique, a patent most likely invalid in light of non-
patent prior art not discovered and consider by the US PTO. Many believe that potential
competitors of Priceline delayed in entering the online auction market while they assessed
(or waited for others to assess) the validity of the Priceline patent. The invalid patent is
a barrier to entry. This is related to the problem European companies will face as they
enter the US software marketplace, when holders of American software patents use their
software patents to delay or prevent the entry of the European company.
2.2.6 Neugr¨ndungen in patentvermintem Gel¨nde
A six set of losers are startups, which sometimes ﬁnd it harder to raise venture capital
if there are many patents, good and bad, in their ﬁeld of software technology. Venture
capitalists see all of these patents as potential mines in the future, so directed their
investment dollars towards startups with lesser degrees of being aﬀected by issued pa-
In short, large numbers of bad software patents distort marketplaces and lead to
transfers of proﬁts not necessarily in the best interests of encouraging and rewarding
innovation. These problems due to American software patents are surely to cross the
Atlantic if Europe liberalizes its software patenting policies.
2.3 Sind die Patent¨mter ihren Aufgaben gewachsen?
With regards to the three major Patent Oﬃces, PTO, EPO and JPO, the answer to this
question is “NO”. The United States Patent and Trademark Oﬃce has been known for
over thirty years to have problems with software patent applications. One of the most
commonly shared complaints is that the PTO for the most part ignores non-patent prior
art, such as conferences and journals, which for software patent applications is often the
most relevant prior art.
For example, 80% of the issued US software patents in the last ten years, eﬀectively cite
no non-patent prior art from any of the world’s engineering and computer societies such
as the IEEE/IEE, ACM and others. Compounding this deﬁciency is the problems the
PTO is having hiring and retaining people with good academic and/or work experience
with software. Patent examiners with bachelors degrees in electrical engineering and a
few years industry experience are not qualiﬁed to examine advanced software patent
Over the last six years, the PTO has made many announcements of new initiatives to
address the problems it is having examining software patents, but it is hard to detect any
improvements in their practices by analyzing recently issued software patents. Recent
court decisions allowing patents on methods of doing business have increased the ﬂood
of software patent applications already overwhelming the PTO, making it likely that the
overload will lead to lower quality software patents from issuing.
Assuming that bureaucracies are globally similar, is not unreasonable that as the
Japanese and European Patent Oﬃces start experiencing levels of software patent app-
lications similar to the 50,000+ software patent applications a year the US Patent Oﬃce
receives, that the JPO and EPO will demonstrate similar levels of problems handling
software patents. Indeed a study I did in 1994 comparing the quality of PCT searches
done by the EPO and PTO found little diﬀerence in the quality, that the EPO searches
were as inadequate as those of the PTO. Assuming that searching conditions haven’t
changed much in recent years at the EPO (especially in light of last falls protests by
EPO examiners about being overworked), then I suspect a study of PCT search quality
for recent PCT applications will ﬁnd similar levels of quality.
So Europe should be careful in allowing expansion of software patenting rights at the
EPO and regional oﬃces, without ﬁrst putting in place some monitoring mechanisms
and support systems to make sure that EPO avoids many of the problems the PTO
demonstrates (partly because there are no independent monitoring mechanisms for PTO
2.4 Verpassen die Europ¨er Chancen?
If one views patents as a form of a trade barrier, then in at least one aspect, European
companies are missing an opportunity. That is, in the United States, which has the
most liberal software patenting policies, American and Japanese companies acquire over
90% of the issued software patents (Americans get about 60%, while the Japanese get
around 30%). Given the comparably sized economics between the United States, Japan
and Europe, one would naively assume that a three way ratio would be (2:1:1), with
Americans getting 46%, Japanese getting 23% and Europeans 25%.
In reality, Europeans are getting about 5% of the US software patents, and thus at a
disadvantage when conducting business in the United States, because European patent
portfolios are too small to trade with American companies such as IBM with very large
portfolios. The cost of entry into the American software market is thus higher, where
European companies aren’t outright prevented from entering the software market due to
blocking patents for which European companies have little to oﬀer other than royalties.
Additionally, if Europe liberalizes its software patenting policies, American and Ja-
panese companies will have an advantage at the outset due to their many decades of
experiences in American and Japan acquiring and asserting software patents. It will
be easy for American and Japanese companies to take many of their patent applicati-
ons already pending and adapt them for European applications (especially those patent
applications pending that have ﬁling date extensions due to PCT protection).