Since the Supreme Court’s Alice¹ decision in 2014, the courts have found many software patents to be invalid as simply being “abstract ideas.” The Patent Office has also been rejecting many new software patent applications as being abstract ideas. Such Patent Office rejections are coming both at the initial examination stage and under various post-grant review systems. Many software patent owners have been feeling rather defeated and frustrated at the odds they have been facing.
A new hope may have arisen. On May 12, 2016, a three-judge Federal Circuit panel offered much needed good news for software patent holders in the case of Enfish LLC v. Microsoft². The issue (as seemingly always for software patents) was whether the claims were patent eligible subject matter under S.101.
The Court held the claims to be patentable subject matter, and this was particularly important since Enfish, the patent applicant, simply claimed a “self-referential” database. There was no physical structure claimed. However, the Court held that this non-physical structure was not simply an abstract idea. The Court in Enfish stated: “[w]e do not read Alice to broadly hold that all improvements in computer-related technology are inherently abstract…”
To date, a common strategy among patent practitioners when writing software patent applications has been to try to claim a physical component of the surrounding computer system or otherwise try to make the claims look as “physical” or as “structural” as possible in an attempt to slip through Alice’s fingers.
Importantly, the Court found that the claims were directed to an improvement to computer functionality versus being directed to an abstract idea. The Court was both influenced by claim language and evidence in the patent’s detailed description that the claimed self-referential table functions differently than a conventional database. The specification described the specific advantages of faster search times, smaller memory requirements and improved flexibility in database configuration. Basically, the Court was convinced that the claimed database structure functioned differently from existing database structures. Moreover, the advantages all related to solving longstanding software problems. Notably, the described advantages were all technical solutions to computer problems.
Therefore, the takeaway from Enfish for patent drafters is that the specification can be as important as the claims. To the greatest degree possible, the specification should explain how the software invention improves the functioning of the overall computer system in which it operates.
The Court also noted that “fundamental economic and business practices are often found to be abstract ideas.” Therefore, it appears to be important to patentablility to not present claims where a general purpose computer is simply being added to a fundamental economic practice or a mathematical equation. The Court had no appetite for these sorts of claims.
Two days later, the Court in TLI Communications LLC v. AV Automotive³ invalidated software claims to a system of organizing digital images. The Enfish case was distinguished with the Court stating that the claims in TLI were not directed to a specific improvement in computer functionality. The Court pointed out that the “specification fails to provide any technical details for the tangible components”. In addition, the Court held that the claims were not directed to a “solution to a technical problem”. Instead, the claims simply described “physical components [that] behave exactly as expected according to their ordinary use.”
On May 19, 2016, the Patent Office issued a Memo4 to its examiners discussing both the Enfish and TLI cases. The Memo stated that the Enfish decision did not change the framework for determining subject matter eligibility, but instead provided “additional clarity” for identifying abstract ideas.
The Memo seemed to raise the bar for discarding software patents as simple abstract ideas. Specifically, it instructed examiners to first identify the abstract idea in the claim and then why it is considered to be abstract. Examiners should also explain their reasoning by reference to previous judicial decisions. The Memo also warned examiners against making sweeping, overbroad statements and requested that they consider the Applicants’ rebuttals.
The Memo urges examiners to look at “the character as a whole” of the claims, as opposed to dissecting the claims. As such, a claim is not doomed simply because it runs on a general-purpose computer. The Memo specifically notes that software can improve computer technology just as much as hardware can. Therefore, an examiner can determine that a claim is not an abstract idea without having to look at additional elements that would confer patentability in the two-part Alice test5. Importantly, the Memo urged examiners to look at the teachings of the specification. It also reminded them that an improvement does not need to be defined by “physical” components. The Memo stated that TLI only referred to “generalized steps to be performed on a computer”, and that is why it was not patentable.
So where does this leave us now?
When writing patent applications, one should focus the claims on the improvements to the computer functionality itself. Both the claims and the specification should preferably be directed to specific improvements to the way computers operate.
Patentable claims are not required to have a physical hardware component. A data structure may be all the “structure” that is required for patentability. That is the good news. However, one should always avoid claims of undue breadth. Also, the writing of the specification was very important in Enfish. A minimalistic specification that merely describes how the claims are enabled, or simply describes what the claims mean is to be avoided. Instead, the specification should be drafted to list advantages of the invention and show how the claimed software structure makes a computer system operate better. The specification is the place to tell a story. Tell a story about the benefits of the invention. Simple claims to business methods performed on computers with the computer operating in an ordinary capacity are still likely to be rejected.