Software Patent Claims

The Alice/Mayo framework is the current test for patent eligibility of software and computer-implemented inventions: ALICE CORP. v. CLS BANK INTERNATIONAL (S.Ct. 2014): holding: claims directed to an abstract idea implemented on a generic computer are patent ineligible under § 101; the Alice test applies the Mayo Collaborative Services v. Prometheus (S.Ct. 2012) two-step framework; STEP 1 — IS THE CLAIM DIRECTED TO A PATENT-INELIGIBLE CONCEPT?: identify what the claims are directed to; if directed to an abstract idea, law of nature, or natural phenomenon → proceed to step 2; if not directed to any ineligible concept → patent eligible (no further analysis needed); WHAT IS AN ABSTRACT IDEA: mathematical concepts: mathematical relationships; mathematical formulas or equations; mathematical calculations; mental processes: concepts performable in the human mind (even with paper and pencil or a computer); certain methods of organizing human activity: fundamental economic principles; commercial interactions; managing relationships between people; 2019 USPTO REVISED GUIDANCE refinement: the USPTO narrowed 'abstract ideas' to three specific groupings above; a claim is NOT directed to an ineligible abstract idea if it does not fall within these three specific categories; STEP 2 — DOES THE CLAIM HAVE AN 'INVENTIVE CONCEPT'?: Step 2A Prong 2: does the claim integrate the abstract idea into a practical application? (positive indicator: technical improvement; specific technical solution; specific machine; transformation); Step 2B: if no practical application, does the claim have additional elements that amount to 'significantly more' than the abstract idea? (conventional computer functions = NOT significantly more; specific non-routine technical steps = possibly significantly more); CURRENT STATE: the Federal Circuit applies Alice strictly; approximately 60-70% of appealed software claims have been found ineligible since Alice; district courts and PTAB also apply Alice to invalidate software patents post-grant.

The 2019 Revised Guidance changed how USPTO examiners apply the Alice framework: THE 2019 USPTO REVISED PATENT SUBJECT MATTER ELIGIBILITY GUIDANCE: issued January 7, 2019; applies to all USPTO examination; courts are NOT bound by this guidance (they continue to apply Alice as interpreted by the Federal Circuit); STEP 2A PRONG 1 (IDENTIFY IF CLAIM RECITES AN ABSTRACT IDEA): the guidance narrowed abstract ideas to THREE specific groupings: (1) mathematical concepts: mathematical relationships; mathematical formulas or equations; mathematical calculations; (2) mental processes: concepts that can be practically performed in the human mind (including with pen and paper; or with a basic computer); (3) certain methods of organizing human activity: fundamental economic principles or practices; commercial or legal interactions; managing personal behavior or relationships; KEY CHANGE: if the claim DOES NOT recite a concept falling within these three groupings, it is NOT directed to an abstract idea → eligible (move to step 1 of the larger framework); this means claims clearly directed to technical processes or systems NOT involving math/mental processes/economic organizing are potentially eligible without further analysis; STEP 2A PRONG 2 (PRACTICAL APPLICATION INTEGRATION): even if the claim recites an abstract idea, it is eligible if the claim as a whole integrates the abstract idea into a PRACTICAL APPLICATION; indicators of practical application: improvement to the functioning of a computer or another technology or technical field; specific machine that is integral to the claim (not incidental); transformation or reduction of a particular article; application of the abstract idea with a particular machine (not a generic computer); STEP 2B (SIGNIFICANTLY MORE): if no practical application → does the claim have additional elements beyond the abstract idea that are significantly more? conventional computer operations = NOT significantly more (Alice holding); non-conventional and non-generic combination of computer components performing non-routine functions = possibly significantly more.

Several claim drafting strategies have proven most effective post-Alice: STRATEGY 1 — TECHNICAL IMPROVEMENT FRAMING: frame the invention as a specific technical improvement to computing technology; not just 'doing X on a computer' but 'improving how computers do X'; EXAMPLES: not 'a method of encrypting data using a computer' but 'a method of reducing memory usage in encryption by performing in-place XOR operations on a circular buffer without additional storage allocation'; the technical improvement is to the computer system itself (memory efficiency; processing speed; throughput; latency); effective technical improvement frames: improved network throughput; reduced latency; reduced memory consumption; improved rendering performance; improved battery life; faster database queries; STRATEGY 2 — SPECIFIC TECHNICAL ELEMENTS: include specific technical components; don't rely on a 'generic processor'; claim specific processors (GPU-based; FPGA-based; neural processing unit); specific memory architectures; specific data structures (hash tables; B-trees; binary space partitions); specific communication protocols; specific hardware interfaces; STRATEGY 3 — ORDERED COMBINATION OF TECHNICAL STEPS: identify the specific ordered combination of technical steps that produces the technical benefit; the combination may be non-obvious even if individual steps are known; specify the sequence and dependency between steps; STRATEGY 4 — INPUT/OUTPUT TRANSFORMATION: focus on a specific transformation of a specific type of data or physical article; a claim that transforms real-world data (sensor data; image data; network packets) into a specific useful result is more likely to be eligible than an abstract manipulation of generic data; STRATEGY 5 — AVOID PURELY FUNCTIONAL LANGUAGE: 'a module configured to analyze data and generate a report' is vague and abstract; 'a processor configured to execute a sliding-window Fourier transform on received audio samples at 44.1 kHz sample rate to identify frequency components above 8 kHz and output a filtered audio stream' is specific and technical.

A comprehensive software patent application includes multiple claim types for maximum coverage: METHOD CLAIMS: 'a method comprising: receiving [specific input]; processing [specific technical operation]; outputting [specific result]'; method claims define the steps of the software process; infringed when someone performs the method; must satisfy § 101 eligibility independently; SYSTEM/APPARATUS CLAIMS: 'a system comprising: at least one processor; memory storing instructions that when executed cause the processor to perform [specific technical operations]'; system claims define the hardware + software combination; same inventive concept as the method but expressed as a structural claim; COMPUTER-READABLE MEDIUM (CRM) CLAIMS: 'a non-transitory computer-readable medium storing instructions that, when executed by a processor, cause the processor to perform [specific operations]'; CRM claims are important for software distributed as code (downloads; apps; firmware); 'non-transitory' is required to exclude transient signals (Beauregard claims); CLAIM HIERARCHY: Claim 1 (independent method): broadest; covers the method at highest level of abstraction consistent with eligibility; Claim 2-5 (dependent method): add specific technical implementations; Claim 6 (independent apparatus/system): same invention in apparatus form; Claims 7-10 (dependent apparatus): technical implementations; Claim 11 (independent CRM): same invention as stored instructions; multiple independent claims ensure that invalidation of one type does not eliminate protection; ALICE CONSISTENCY: all three claim types must independently satisfy § 101; if the method claim recites the technical improvement, the system and CRM claims must also capture the technical improvement (not just 'a system that performs the method' in the abstract); AI AND MACHINE LEARNING CLAIMS: claim the specific training method (training data types; loss function; optimization algorithm); claim the specific inference architecture (layer types; attention mechanisms; specific transformations); avoid claiming 'using AI' without technical specificity — that is abstract.

Responding to § 101 rejections requires both legal arguments and potential claim amendments: ANALYZING THE REJECTION: identify which step the examiner rejected at (Step 2A Prong 1; Prong 2; or Step 2B); understand which abstract idea category the examiner identified (mathematical concept; mental process; economic organizing); determine whether the examiner applied the 2019 Revised Guidance correctly; ARGUMENT-ONLY RESPONSE (preferred if possible): argue that the claims do NOT recite the identified abstract idea OR that they integrate the abstract idea into a practical application; PRONG 1 ARGUMENT: the claim does not fall within any of the three abstract idea groupings; it recites a specific technical process or system, not a mathematical concept, mental process, or method of organizing human activity; PRONG 2 ARGUMENT: even if abstract, the claim integrates the abstract idea into a practical application because [describe the specific technical improvement; specific machine; transformation]; cite MPEP 2106.05(a) improvements to technology; cite Federal Circuit cases with favorable outcomes: Enfish v. Microsoft (self-referential table improves database performance); McRO v. Bandai Namco (specific rules for character lip-sync); Core Wireless v. LG Electronics (specific interface for small-screen applications); CLAIM AMENDMENT APPROACH (if arguments insufficient): add specific technical elements that clearly remove the claim from abstract idea territory; add the specific technical improvement language ('wherein the method reduces cache misses by...'); add specific hardware elements ('wherein the processor is an application-specific integrated circuit (ASIC) configured to...'); EXAMINER INTERVIEW: request an examiner interview before filing a response to align on what specific technical elements would overcome the rejection; this saves amendment cycles; AFTER-FINAL OPTIONS: file an RCE after a final rejection to re-open prosecution; consider appeal to PTAB if arguments are strong and examiner is incorrectly applying the 2019 Revised Guidance; include declaration evidence (§ 1.132) about the technical improvement if commercial or technical evidence supports it.