Computer-Related Inventions (CRIs) are inventions that involve computer programs as part of the claimed process or apparatus. While "computer programmes per se" are excluded from patentability in India (S.3(k)), inventions with a technical contribution beyond the program itself may qualify for patent protection.
Legal Framework
| Provision | Subject |
|---|---|
| Patents Act S.2(1)(j) | "Invention" = new product or process involving inventive step, capable of industrial application |
| Patents Act S.2(1)(ja) | "Inventive step" = feature not obvious to person skilled in art, involving technical advance or economic significance |
| Patents Act S.3(k) | Exclusion: mathematical method, business method, computer programme per se, algorithms |
| Patents Act S.2(1)(ac) | "Capable of industrial application" = can be made or used in industry |
| CRI Guidelines 2017 (CGPDTM) | Examination procedure for computer-related inventions |
| EPC Art.52(2)(c) | Programs for computers excluded "as such" |
| 35 USC 101 (US) | Patentable subject matter (utility patents) |
Indian Position: Section 3(k) and CRI Guidelines
The "Per Se" Qualifier
S.3(k) excludes "a computer programme per se." The words "per se" limit the exclusion:
| Claim Type | Patentable? | Reasoning |
|---|---|---|
| Claim ONLY for software code | No | Computer programme per se |
| Claim for hardware controlled by software | Yes | Technical apparatus with software component |
| Claim for technical process using algorithm | Potentially yes | Technical effect beyond mere computation |
| Claim for business method on computer | No | Business method excluded regardless of implementation |
| Claim for improved computer performance | Potentially yes | Technical improvement in computer's own functioning |
| Claim for data compression algorithm alone | No | Algorithm per se |
| Claim for device using novel data compression | Potentially yes | Technical effect (reduced bandwidth/storage) |
CRI Guidelines 2017 (CGPDTM)
Three-step examination approach:
- Identify the actual contribution of the claimed invention (look beyond claim language to technical substance)
- Determine if contribution lies solely in a computer program (if yes, reject under S.3(k))
- If technical effect exists, assess novelty and inventive step as normal
Why: India takes a middle path: not as liberal as the US (pre-Alice) nor as restrictive as a literal reading of S.3(k) might suggest. The "technical effect" test ensures genuinely innovative technical solutions using software can still be patented.
European Position
| Patentable | Not Patentable |
|---|---|
| Software producing "further technical effect" | Software "as such" (programs that merely instruct a computer) |
| Computer-controlled manufacturing process | Business methods implemented on computer |
| Medical device software | Pure mathematical methods |
| Improved telecommunications method | Presentation of information |
| Data processing with real-world technical application | Aesthetic creations |
EPO "Problem-Solution" approach: Does the software solve a technical problem in a non-obvious way? If yes, patentable regardless of software involvement.
US Position (Post-Alice)
Alice/Mayo Framework (Two-Step Test):
| Step | Question | If Yes |
|---|---|---|
| Step 1 | Is the claim directed to a patent-ineligible concept (abstract idea, law of nature)? | Proceed to Step 2 |
| Step 2A | Does the claim recite additional elements amounting to "significantly more" than the abstract idea? | Patentable |
| Step 2B | Or is it merely conventional computer implementation? | Not patentable |
Post-Alice, thousands of US software patents have been invalidated. The test has made software patentability substantially harder in the US.
Standard Essential Patents (SEPs) in Technology
| Concept | Meaning |
|---|---|
| SEP | Patent essential to implement an industry standard (e.g., 4G/5G, Wi-Fi, Bluetooth) |
| FRAND | Fair, Reasonable, and Non-Discriminatory licensing terms for SEPs |
| Patent holdup | SEP holder demands excessive royalties knowing implementer is locked into standard |
| Patent holdout | Implementer refuses to negotiate FRAND license |
| Indian position | Ericsson v. Intex (2015): SEP holders must offer FRAND terms; injunction possible for unwilling licensees |
Recall Check
- What is the three-step CRI examination approach under 2017 Guidelines?
- What distinguishes a "computer programme per se" from a patentable CRI?
- What is the Alice two-step test?
Key Cases
Diamond v. Diehr (1981) Diamond-v-Diehr-1981 Issue: Whether a rubber-curing process using a computer program to calculate cure time is patentable. Rule: A claim to an industrial process is not rendered unpatentable merely because one step uses a mathematical formula implemented by a computer. Held: US Supreme Court held the process patentable. The claim was for a physical process (curing rubber), not for the algorithm itself.
Ferid Allani v. Union of India (2020) Ferid-Allani-v-Union-of-India-2020 Issue: Whether patent office can refuse a CRI application solely on the ground that it involves a computer program. Rule: "Per se" in S.3(k) limits the exclusion; inventions with technical effect beyond the program itself cannot be mechanically rejected. Held: Delhi HC remanded for fresh consideration. Patent office must assess technical contribution, not mechanically apply S.3(k).
Telefonaktiebolaget LM Ericsson v. Intex Technologies (2015) Ericsson-v-Intex-Technologies-2015 Issue: Whether a SEP holder can seek injunction for infringement when the implementer refuses to negotiate FRAND license. Rule: SEP holders must offer FRAND terms before seeking injunction; implementers must negotiate in good faith. Unwilling licensees may face injunction. Held: Delhi HC granted interim injunction. Ericsson demonstrated willingness to license on FRAND terms; Intex was an unwilling licensee. Established Indian framework for SEP enforcement.
Distinctions
| Aspect | Computer Programme Per Se (Not Patentable) | Computer-Related Invention (Potentially Patentable) |
|---|---|---|
| Claims directed to | Code, algorithm, method of operation | Technical process, apparatus, device using software |
| Technical effect | None beyond running on computer | Solves technical problem, improves performance |
| Example | Algorithm for sorting data | Device using novel sorting to reduce memory usage by 50% |
| S.3(k) | Squarely excluded | May qualify if technical contribution shown |
| Drafting focus | Steps of computation | Technical problem + technical solution + technical result |
| Indian authority | Rejected by Patent Office | Ferid Allani (remanded for fresh examination) |
Flashcards
Q: What does S.3(k) Patents Act exclude? A: "A mathematical or business method or a computer programme per se or algorithms."
Q: What is the significance of "per se" in S.3(k)? A: Limits the exclusion to programs in isolation; inventions with technical effect beyond the program itself may still be patentable.
Q: What is the CRI Guidelines 2017 three-step test? A: (1) Identify actual contribution of invention, (2) Determine if contribution lies solely in computer program, (3) If technical effect exists, assess novelty and inventive step normally.
Q: What are FRAND terms? A: Fair, Reasonable, and Non-Discriminatory licensing terms that holders of Standard Essential Patents must offer to implementers of the standard.
Q: What did Ericsson v. Intex (2015) establish? A: Indian framework for SEP enforcement: SEP holder must offer FRAND terms; unwilling licensees can face injunction; willing licensees cannot be injuncted.
Q: What is the EPO "further technical effect" test? A: Software is patentable if it produces a technical effect beyond the normal physical interaction between program and hardware (e.g., improved efficiency, solving specific technical problem).
Q: Can a business method implemented on a computer be patented in India? A: No. S.3(k) excludes both "business methods" and "computer programmes per se." A business method does not become patentable merely by computerizing it.
Exam Scenario
A pharmaceutical company develops a novel AI-based system that predicts drug interactions with 95% accuracy using a proprietary machine learning algorithm. They file a patent in India. The Patent Office rejects under S.3(k) stating it is a "computer programme per se." Advise.
The rejection is challengeable under Ferid Allani (2020). The AI system must be analyzed for technical contribution: (1) It solves a technical problem (predicting drug interactions with quantifiable accuracy improvement), (2) The contribution is not merely the algorithm but the trained model + database + novel methodology producing medically significant results, (3) It has industrial application (pharmaceutical development, clinical safety). The claim should be redrafted (if necessary) to emphasize: (a) the technical problem (drug interaction prediction failures), (b) technical means (novel ML architecture + pharmaceutical database integration), (c) technical result (95% accuracy, superior to existing methods). Under CRI Guidelines 2017, the examiner must assess whether the contribution lies solely in the computer program. Here, the contribution includes: novel training methodology, domain-specific database application, and measurable pharmaceutical safety outcome. The claim should survive S.3(k) because the invention produces a "further technical effect" in the pharmaceutical domain.