Abstract Idea Test

The § 101 framework tests whether a patent claim is directed to patent-eligible subject matter (process, machine, manufacture, or composition of matter) or to an ineligible exception (law of nature, natural phenomenon, or abstract idea): ALICE CORP. v. CLS BANK INTERNATIONAL (S.Ct. 2014): established the two-step framework for patent eligibility of software and business method claims; extended Mayo Collaborative Services v. Prometheus Labs (S.Ct. 2012) to non-biologic claims; STEP 1 — DIRECTED TO AN ABSTRACT IDEA?: does the claim, when considered as a whole, recite an abstract idea?; examples of abstract ideas: mathematical concepts (formulas, algorithms); fundamental economic concepts (hedging risk, intermediated settlement); methods of organizing human activity (a game, a business method); mental steps that a human could perform in their head; IF NO ABSTRACT IDEA: the claim is eligible — no further analysis needed; most modern software claims fail Step 1 (they are found to recite abstract ideas); STEP 2 — DOES IT ADD SIGNIFICANTLY MORE? (if Step 1 = yes): does the claim recite additional elements that transform the abstract idea into a patent-eligible application?; SIGNIFICANTLY MORE must be beyond the abstract idea itself; a specific technical improvement to computer technology (not just using a computer); NOT significantly more: applying the abstract idea on a generic computer; applying the abstract idea using conventional software steps; limiting to a particular field without changing the abstraction; USE OF A COMPUTER ALONE IS NOT ENOUGH: 'apply it with a computer' does not transform an abstract idea into eligible subject matter; the computer must be doing something new or solving a technical problem in a new way.

The USPTO issued Revised Patent Subject Matter Eligibility Guidance in January 2019, refining how examiners apply Alice: 2019 PEG STEP 2A PRONG 1 — WHAT IS THE ABSTRACT IDEA?: examiners must identify the specific abstract idea by reference to three groupings: (a) MATHEMATICAL CONCEPTS: mathematical relationships, mathematical formulas or equations, mathematical calculations; (b) MENTAL PROCESSES: concepts that humans perform in their minds or with pen and paper; (c) CERTAIN METHODS OF ORGANIZING HUMAN ACTIVITY: fundamental economic principles, commercial interactions, managing personal behavior or relationships; IMPORTANT LIMITATION: an examiner cannot simply call something an 'abstract idea' without mapping it to one of these three groupings; vague assertions that a concept is 'abstract' without specificity are improper rejections under the 2019 PEG; 2019 PEG STEP 2A PRONG 2 — PRACTICAL APPLICATION?: does the claim as a whole integrate the abstract idea into a practical application?; this analysis considers: (a) whether the additional elements reflect a specific technological improvement (not just the abstract idea applied to a computer); (b) whether the claim applies a mathematical concept to a particular article or machine; (c) whether the claim uses a particular machine or transformation (though this is not required post-Bilski); IF PRACTICAL APPLICATION = YES: the claim is eligible without reaching Step 2B; STEP 2B — WELL-UNDERSTOOD, ROUTINE, CONVENTIONAL?: only reached if there is no practical application identified; are the additional elements beyond the abstract idea well-understood, routine, conventional (WURC) in the relevant field?; if the additional elements are WURC, the claim is ineligible; EXAMINER BURDEN: examiners must provide evidence (prior art, admitted prior art, official notice) that the additional elements are WURC; they cannot simply assert it.

Courts and the USPTO have established patterns for what survives and fails the abstract idea test: CLAIMS THAT TYPICALLY SURVIVE: (a) TECHNICAL IMPROVEMENTS: claims that address a specific technical problem in a computer or network in a new way (improved data compression, reduced processing latency, improved memory efficiency); Enfish v. Microsoft (Fed. Cir. 2016): self-referential table database structure improved computer memory — eligible; (b) SPECIFIC HARDWARE COMBINATION: claims that tie the abstract idea to a specific hardware architecture beyond a generic computer; (c) UNCONVENTIONAL COMPUTER OPERATIONS: claims where the specific computer operations are themselves the improvement, not just the result; CLAIMS THAT TYPICALLY FAIL: (a) GENERIC COMPUTER + ABSTRACT IDEA: any mathematical or business method claim that simply says 'perform this on a computer' without a specific technical improvement; (b) FINANCIAL AND COMMERCIAL METHODS: hedging, intermediated settlement, risk mitigation, bilateral trading — all abstract; Alice itself: computerized method for mitigating financial settlement risk using escrow — abstract; (c) DATA MANIPULATION WITHOUT TECHNICAL IMPROVEMENT: collecting data, analyzing data, displaying results — abstract unless the collection, analysis, or display itself is technically improved; Elec. Power Grp. v. Alstom (Fed. Cir. 2016): monitoring and analyzing power grid data — abstract; (d) MENTAL STEPS + COMPUTER: anything a human could do in their head and a computer just does faster; (e) AI/MACHINE LEARNING CLAIMS: training a neural network using conventional backpropagation + standard training data — likely abstract; applying a model to generate predictions — likely abstract; the specific ML architecture or algorithm must be technically novel; MOST VULNERABLE AREAS: fintech, business software, data analytics, ad tech, communication methods.

Section 101 (patent eligibility) and § 102/103 (novelty and non-obviousness) serve different functions in the patent system: § 101 — THRESHOLD ELIGIBILITY: determines whether the invention belongs to a patentable CATEGORY at all; a claim that fails § 101 cannot be patented regardless of how novel or non-obvious it is; it is a threshold question, not a patentability merits question; does NOT require comparison to prior art — a claim can be brand-new AND still fail § 101; § 102/103 — MERIT-BASED PATENTABILITY: § 102 novelty: does prior art anticipate the claim (is every element present in a single prior art reference)?; § 103 non-obviousness: would the claim have been obvious to a POSITA given the prior art?; THESE REQUIRE PRIOR ART COMPARISON: the claims are evaluated against the prior art; a truly novel, non-obvious claim still survives § 102/103 even if the prior art is close; CONFUSION BETWEEN THEM: sometimes examiners conflate § 101 'abstract idea' with § 103 obviousness; the two are distinct; a claim can be both obvious (§ 103) AND directed to an abstract idea (§ 101) — but each requires separate analysis; courts have been criticized for using § 101 to invalidate claims that could have been analyzed under § 102/103: 'the concern is that § 101 will be used to invalidate patents that merely lack novelty or are obvious' (Berkeley Chem Sec.); PRACTICAL OVERLAP: 'well-understood, routine, and conventional' in Step 2B of Alice sometimes overlaps with 'obvious' under § 103; but WURC in § 101 is a narrower inquiry — it asks whether the additional elements (not the whole claim) are conventional; § 103 asks whether the whole claim would have been obvious to try or combine; PROSECUTION STRATEGY: pursue both § 101 and § 102/103 responses simultaneously; do not concede § 101 eligibility of the prior art claims while arguing § 102/103.

Strategies for drafting § 101-eligible claims focus on grounding abstract ideas in specific technical implementations: STRATEGY 1 — CLAIM THE TECHNICAL IMPROVEMENT: identify the specific technical problem the invention solves; claim the solution in terms of what the computer system DOES differently, not what result it achieves; instead of 'a method for recommending products,' claim 'a method for reducing cache collisions in a distributed recommendation system by [specific technique]'; STRATEGY 2 — INCLUDE SPECIFIC HARDWARE OR NETWORK ARCHITECTURE: tie the claim to specific hardware or network components beyond a generic processor; 'a network switch configured to [specific novel function]' is better than 'a processor configured to [abstract function]'; STRATEGY 3 — CLAIM THE NOVEL ALGORITHM SPECIFICALLY: if the core innovation is an algorithm, claim it in detail; do not describe results — describe the specific computational steps; include the technical reason the algorithm is better (faster, more accurate, more efficient in a measurable technical way); STRATEGY 4 — LEAD WITH THE TECHNICAL PROBLEM IN THE PREAMBLE: 'A computer-implemented method for reducing latency in [technical system] comprising...' situates the claim in a technical context; STRATEGY 5 — AVOID RESULT-ORIENTED CLAIMING: 'a method for generating accurate predictions' is a result; 'a method comprising: training a model using [specific technique], applying [specific loss function], achieving [specific technical parameter improvement]' is a technical implementation; STRATEGY 6 — SPECIFICATION MATTERS: write detailed specification sections explaining the specific technical problem and why conventional approaches fail; this context helps claim construction and § 101 analysis; cite technical benchmarks showing improvement; STRATEGY 7 — DIVISIONAL AND CONTINUATION STRATEGY: if a broad method claim is vulnerable to § 101, file narrower claims focused on specific technical embodiments as continuation or divisional claims; STRATEGY 8 — BIFURCATE METHOD AND SYSTEM CLAIMS: method claims for § 101 face higher scrutiny than system (apparatus) claims in some courts; draft both to maximize eligibility arguments.