Abstract Ideas

Abstract ideas are one of three judicial exceptions to patent eligibility established by the Supreme Court: THE THREE JUDICIAL EXCEPTIONS: under 35 U.S.C. § 101, four categories of subject matter are potentially patentable: processes, machines, manufactures, and compositions of matter; but courts have carved out three judicial exceptions: (1) laws of nature; (2) natural phenomena; (3) abstract ideas; ABSTRACT IDEAS ARE NOT PATENTABLE BECAUSE: granting a patent on an abstract idea would 'preempt' all uses of that idea — the patent would cover every application of the concept regardless of how it is implemented; the Constitution's patent clause intends to promote 'Progress of Useful Arts' — pure abstract concepts do not represent a technological contribution; THE DEFINITION PROBLEM: neither Congress nor the Supreme Court has precisely defined 'abstract idea'; the Supreme Court has said: 'we need not define... the precise contours of the 'abstract ideas'... We need only recognize that there is no dispute that [the Alice patents were] directed to an abstract concept' (Alice, 2014); the lack of a precise definition makes § 101 analysis unpredictable; HOW THE USPTO DEFINES ABSTRACT IDEAS (2019 Revised Guidance): three groupings: (a) Mathematical concepts: mathematical relationships; mathematical formulas or equations; mathematical calculations; (b) Certain methods of organizing human activity: fundamental economic principles or practices (hedging; intermediated settlement; risk management); commercial or legal interactions (contracts; sale/purchase; business relations); managing personal behavior or relationships (following rules or instructions; human activity coordination); (c) Mental processes: concepts performed in the human mind (observation; evaluation; judgment; opinion); even if a computer is used for speed/accuracy, if a human COULD perform the process mentally, it may be a mental process; WHY IT MATTERS: an estimated 30-40% of all software patent claims filed between 2000 and 2014 were potentially vulnerable to Alice/Mayo rejection; since Alice, the USPTO issued thousands of § 101 rejections and courts invalidated many issued patents.

Certain technical fields are particularly susceptible to abstract idea rejections: FINTECH/FINANCIAL SERVICES: fundamental economic practices: hedging (Bilski v. Kappos, S.Ct. 2010: method for hedging commodity risk = abstract idea); intermediated settlement (Alice v. CLS Bank, 2014); risk management and mitigation; economic calculations: interest calculation methods; tax reporting; currency conversion algorithms; THE PATTERN FOR FINANCIAL PATENTS: a financial concept + implement it on a computer = abstract idea + apply it = ineligible; must have something more — a specific technical improvement to computer infrastructure (not just applying the concept faster); SOFTWARE AND APP DEVELOPMENT: display/organization of information: displaying real estate listings in a certain order; organizing/filtering content; interface design without a specific technical improvement; data processing: collecting user data; analyzing user behavior; matching users with items; recommendations; the pattern: routine data + analyze it + present results = abstract idea pipeline; ARTIFICIAL INTELLIGENCE/MACHINE LEARNING: certain ML-based patents face abstract idea challenges; a claim to 'apply machine learning to predict X' without a specific technical improvement may be ineligible; a claim to a specific neural network architecture that improves processing efficiency is more defensible; INTERNET AND ECOMMERCE: online shopping cart (collecting order information); subscription management; targeted advertising (correlating user data with advertisement content); access control and authentication (verifying identity using abstract steps); HEALTHCARE IT: storing and retrieving patient health records; correlating health data to risk scores; health data presentation and dashboards — without specific technical improvements to computer systems; WHAT MAKES THESE ABSTRACT: they describe WHAT is accomplished (collect data, analyze, present), not HOW the computer accomplishes it in a technically non-obvious way.

The line between abstract and eligible is determined by whether the claim reflects a specific technical contribution: ENFISH PRINCIPLE (Fed. Cir. 2016): a software claim is eligible if it is directed to a specific improvement in computer functionality itself; not all improvements using a computer — but improvements TO computer functionality; Enfish LLC v. Microsoft: a self-referential database table that stored all data in a single table (including metadata) was found eligible because it improved the efficiency of the database (faster searching, less memory) compared to conventional relational database tables; it was a TECHNICAL improvement to how computer memory and data structures work; MCRO PRINCIPLE (Fed. Cir. 2016): a specific non-conventional ordered combination of computer steps may be eligible even if individual steps are abstract; McRO v. Bandai Namco: 3D lip-sync animation automation using specific rules for transforming text input into morph weights; the specific unconventional ordered combination of steps to produce a realistic animation was eligible; the claim was to a specific way of doing the transformation, not just 'automate animation on a computer'; CORE WIRELESS (Fed. Cir. 2018): a specific improved user interface for mobile devices was eligible; the specific arrangement of application data in a summary window accessible from a main menu — an improved user interface design with a specific technical structure; KEY QUESTION: is the claim directed to a specific technical implementation (specific structures, steps, arrangements) that improves computer functionality or creates a new computer capability? OR is it directed to a concept that uses a computer as a tool to implement something fundamentally non-technical?; USEFUL HEURISTIC: replace 'computer' with 'human with a pen and paper' — if the claim still works conceptually, it may be an abstract mental process + computer; if it cannot work without specific computer functionality, it may be eligible.

Mental process claims are those that can be performed by a human mind, even if assisted by a computer: MENTAL PROCESS CASES: CyberSource Corp. v. Retail Decisions (Fed. Cir. 2011): method for detecting credit card fraud by mapping transaction relationships to trace a card number back to other transactions; held a mental process because a human could review transaction records and detect the patterns without a computer; the computer just makes it faster; Classen Immunotherapeutics v. Biogen (Fed. Cir. 2011): method of comparing immunization schedules based on comparing group health outcomes; abstract mental comparison without physically transforming anything; FairWarning IP v. Iatric Systems (Fed. Cir. 2016): software for detecting healthcare privacy violations by analyzing audit logs; held abstract mental process: analyzing and comparing log data for suspicious patterns is a mental process humans could do manually; ELIGIBLE COUNTEREXAMPLES: SRI International v. Cisco Systems (Fed. Cir. 2019): detecting network intrusions by combining data from hierarchical network monitors; eligible because: the claim was to a specific technical architecture (hierarchical monitors); the combination of data from different monitoring levels was technically non-conventional; not merely 'observe + compare + flag'; Improved performance of network security systems = technical improvement; THE MENTAL PROCESS TEST: can a POSITA mentally perform the claimed process (even if slowly or impractically)?; 'using a computer speeds up' the mental process is insufficient — speed/accuracy on a computer does not transform a mental process; 'a computer is the only practical way to do this' + specific technical steps that are not just 'apply it' = potentially eligible; PRACTICAL DRAFTING: avoid pure 'collect-analyze-present' claim structures; include specific data structures, specific algorithms, specific hardware interactions; describe what makes the computational approach technically different from what could be done mentally or conventionally.

Practitioners have developed specific techniques for improving § 101 defensibility: CLAIM DRAFTING STRATEGIES: ANCHOR TO A TECHNICAL PROBLEM: identify a specific technical problem (memory overflow; latency bottleneck; computational inefficiency; security vulnerability); the claim solves that specific technical problem with a specific technical solution; avoid framing the invention as solving a business problem implemented on a computer; SPECIFIC TECHNICAL ELEMENTS: recite specific non-generic computer components; don't say 'a processor' — say 'a processor configured to execute [specific algorithm] using [specific data structure]'; describe specific hardware-software interactions; include specific parameters, specific algorithms, specific configurations; ORDERED COMBINATION: even if each individual step is conventional, an unconventional ordered combination of steps that produces an unexpected technical result may be eligible; describe WHY the specific ordering matters technically; TREATMENT/TRANSFORMATION: for diagnostic applications, anchor to a treatment step with specific drug, dose, and patient characteristics (Vanda model); for physical processes, describe the physical transformation (transformation and machine test); SPECIFICATION STRATEGIES: include technical problem + solution description explicitly at the beginning of the description; include performance data showing technical improvements (speed, accuracy, memory, security) achieved by the specific technical approach; include drawings showing specific computer architecture, data flows, or novel hardware configurations; PROSECUTION STRATEGIES: in response to § 101 rejection, argue the specific practical application of the judicial exception (2019 Guidance Step 2A Prong 2); argue technical improvements using specific cases (Enfish, McRO, Core Wireless) as analogies; avoid purely functional claiming ('wherein the processor performs function X') without structural support; narrowing claims to specific technical implementations is often more productive than fighting the abstract idea characterization.