As followers of IP would attest, IP law develops in two ways, firstly via new legislation, and secondly through new (or revised) interpretations of existing laws, applied during patent law cases that can end up the highest courts in the land - and hence become influential on lower courts who are required to apply these new interpretation whenever applicable.
Both processes are ongoing, with new patent legislation before the US Senate aimed at dealing with so-called patent trolls, and with the recently announced US Federal Circuit decision in CLS Bank v. Alice Corp making patent professionals think again about software patents.
Each of the cases that lead to revised interpretations are based on real court cases, and often in relation to questions of validity of patents that challenge the existing understanding of the limits of patentable subject matter.
So what are the types of patent that can change patent law? Do they deserve their exulted status?
There are many candidate for such patents, but patents for five of the most relevant cases that are still impacting on modern patent law are reviewed below. In each review we will concentrate on the patents and not the legal aspects of the case - these have often been well covered by legal scholars.
1) Diamond v. Chakrabarty (1980) concerned the question of whether a newly developed bacterium capable of breaking down crude oil could be patented by General Electric, the employer of Chakrabarty. The examiner rejected the patent on the grounds that he understood at the time that living things could not be patented. However the US Supreme Court allowed the patent, ruling that patentable subject matter to 'include anything under the sun that is made by man'. This definition has since been used to justify the patenting of a wide variety of inventions, including business methods and software.
The patent in question is US 4,259,444, filed in 1972. When reviewed in AmberScope as shown below, it can be seen that while this patent had a huge impact in patent law, it only had a mild impact on the patent literature. The '444 patent has an AmberScore value of 0.19, or 19% of the average for US granted patents (admittedly filed in the last 20 years). This may reflect its zero backward citations, and just 21 forward citation in over 40 years.
Another interesting test of relative quality is to compare the AmberScore value of the patent being reviewed to the AmberScore values of its directly connected neighbors (forward and backward citations). The median AmberScore value for the patents the '444 patent is directly connected to is 0.61 - and in fact, this patent sits in the bottom 20% of these patents when rated by AmberScore value.
So for this patent, its impact on the patent literature has been smaller than its influence on patent law.
2) State Street Bank v. Signature Financial Group (1998) concerned a case where the US Federal Circuit held that a patent in relation to setting a share price to minimise tax was patentable as it "it produces a useful, concrete and tangible result" - . In doing so, the Federal Circuit essentially confirmed the patentability of business patents, a decision that is thought to have been overturned by the Supreme Court ruling on the validity of the next patent on this list.
But returning to State Street, the patent in question was U.S. Patent 5,193,056, filed in 1993. In contrast to the Chakrabarty patent, the Signature Financial Group owned patent has had a strong impact on the patent literature, with an AmberScore value of 6.7. This includes contributions from 277 forward citations. Similarly, an AmberScore value of 6.7 compares well to a median score for 3.3 for its directly connected patents, and sits in the top 25% of these patents.
The intensity of the connection to the Signature patent is shown in the image below, which shows a crowded patent network.
3) Bilski vs Kappos (2010). State Street may have opened up the door for business method patents, but Bilski is thought to have started the process of closing this door. First filed in 2002, Bilski claimed a method of hedging risks in commodities trading via a fixed bill system. Rejected by the USPTO, Bilski was litigated all of the way up the US Supreme Court, who rejected the patent in 2010 because the claimed invention did not "transform any article to a different state or thing". While legal obligations were transformed by this invention, they "cannot meet the test because they are not physical objects or substances, and they are not representative of physical objects or substances."
The patent application in question was US20040122764. Besides being unsuccessful in attracting patent protection, the Bilski patent was also unsuccessful in attracting citation connections. The image below shows this the Bilski patent, with an AmberScore value of just 0.01, or 1% of the average AmberScore value for granted US patents less than 20 years old. All of the connected patents have higher AmberScore values, but with a low median value of 0.42.
4) Mayo v. Prometheus (2012). The next major narrowing of patentable subject matter came in 2012 with Mayo v. Prometheus. This case concerned an invention for a feedback loop that adjusted the dosage of a drug depending the measured response of a body to the drug. The US Supreme Court ultimately rejected the patent, on the grounds that it was an unpatentable "natural law" and not “genuine applications of those laws ....rather ... drafting efforts designed to monopolize the correlations."
There were two patents in question, namely US 6,355,623 (AmberScore value 1.1, 3 forward citations, 2 backward citations) and US 6,680,302 (AmberScore value 1.3, 1 forward citation, 5 backward citations). The AmberScore map for the higher scoring 302 patent is shown below (besides the directly connected patent, this map shows potentially relevant indirectly connected patents.
The median AmberScore values for the directly connected patents was 0.3, with the the '302 patent having a higher AmberScore value than any of these patents, including the '623 patent.
5) CLS Bank v. Alice Corp (2013). The last patent is a recent Federal Circuit en banc ruling that may be appealed to the Supreme Court, but already has thought to have wide ranging implications. CLS Bank concerns an application for four patents, which including claims for a business method of reducing counter-party risk during trading using a third party - and claims for the application of this business method using a computer. While the different judges on the panel gave slightly different opinions, the dominant opinion can be simplified to
- The business method claims were abstract ideas and therefor not patentable
- And more controversially, incorporating the use of a computer into a similar claim was 'insignificant post-solution activity' because 'Abstract methods do not become patent-eligible machines by being clothed in computer language'.
The four patents in question were:
- US 5,970,479 filed in 1993, (AmberScore value of 23, 234 forward citations, 30 backward citations)
- US 6,912,510, filed in 2000 (AmberScore value of 6.7, 28 forward citations, 61 backward citations)
- US 7,149,720, filed in 2002 (AmberScore value of 5.3, 2 forward citations, 63 backward citations)
- US 7,725,375, filed 2005 (AmberScore value of 5.6, 13 forward citations, 92 backward citations)
So all patents have respectable AmberScore values. This is demonstrated in the very crowded patent network for the first listed of these patents, shown below. The median AmberScore value of the directly connected patents is 1.73, and the AmberScore value of the '470 patent was enough to put this into the top 10% of this crowded patent landscape.
When comparing these patents, three of these patents had comparatively low AmberScore values, showing in this limited sample a non-existent relationship between high AmberScore values (and forward citation count) and litigation. Does this mean that such measures are a poor predictor of important patents, since litigated patents are more likely to be important than non-litigated patents?
Not at all. Without doubt, litigated patents are important to their owners, but patent owners are not always a truly objective judge of the value of a patent. More to the point, for the validity of a patent to be litigated up to superior courts requires an opponent with a belief that the validity of the patent is arguable, and be prepared to spend a significant amount of money to test this argument. It also requires courts that are willing to hear these cases, that is willing to accept that there is an argument to be heard.
Imagine, in contrast a hypothetical high quality patent that an opponent reviews and finds no substantial grounds to oppose. They would be more likely to settle a dispute in relation to this patent than risk the cost, bad publicity, embarrassment, and court ordered damages that could result from what would be likely to be a unsuccessful attempt to invalidate the patent.
For this reason, Ambercite, unlike some other analysts, does not include litigation history when assessing patent quality - because many of the best patents may never be litigated.
PS - A future candidate for the list of patents that may change patent law?
I have already commented twice on the 'using a scanner to email documents' patent, a patent that is still causing problems to small business owners in the US. An earlier post pointed out how relatively weak this patent was according to an AmberScore analysis. Patents such as this have appeared to catalyze new legislation aimed at so called 'patent trolls' - if this legislation is passed, this particular patent is sure to be mentioned as a reason.
The patent network for this patent is shown here - and confirms that this is a low (predicted) value patent. An AmberScore value of 0.03 places this patent dead last last among its directly connected patents.