India’s Draft CRI Guidelines 2.0 (2025): A Sharper Blueprint for Patentability in the Age of AI
- IPR IN INDIA
- Jul 5
- 5 min read
Issued on June 26, 2025, the Indian Patent Office’s revised Draft Guidelines for Examination of Computer-Related Inventions (CRI) marks a significant step forward in shaping a more predictable and robust framework for software and AI-related patents in India. With public consultation open until July 7, 2025, stakeholders now have an opportunity to help shape a document that will directly impact the way emerging technologies - ranging from AI and blockchain to quantum systems - are evaluated for patent protection.
From Draft to Draft 2.0: Why the Update Matters:
This revised version of the guidelines - referred to informally as CRI Guidelines 2.0 - follows an earlier draft published in March 2025. The update was prompted by extensive feedback from industry, academia, and legal professionals, all of whom raised concerns around clarity, applicability to emerging technologies, and inconsistency in examination practice.
Version 2.0 retains the spirit of the original but improves upon it with:
§ Expanded illustrative examples, especially for AI and cryptography,
§ Better integration of judicial precedent, and
§ A more structured approach to assessing novelty, inventive step, and technical effect.
What Has Changed: Key Highlights from CRI 2.0
1. Emphasis on Technical Effect and Contribution
The core test for patent eligibility remains unchanged: computer-related inventions must produce a technical effect. What’s improved is the depth of examples provided to illustrate what this means in real terms - such as improved signal processing, real-time robotic control, and enhanced cybersecurity protocols.
2. Judicial Precedents Anchored into the Text
The revised guidelines ground the eligibility analysis in Indian case law, including Ferid Allani, Microsoft Technology Licensing, and OpenTV. This move provides much-needed alignment between examination standards and judicial reasoning, enhancing both predictability and legitimacy.
3. Structured Examination: “Seven Stambhas” & Five-Step Tests
To assess novelty, the guidelines adopt the “Seven Stambhas” approach from Ericsson v. Lava. For inventive step, they reaffirm the classic five-step analysis rooted in Indian Supreme Court jurisprudence.
Annexure I: The New Illustrative Examples
Perhaps the most visible improvement in CRI 2.0 is the inclusion of Annexure I, which provides over 20 examples - both patent-eligible and non-eligible. These cover AI-based systems, blockchain innovations, and exclusions such as data sorting methods or business automation tools lacking technical novelty.
Disclosure Requirements: Higher but Clearer Bar
The guidelines reinforce the need for sufficiency of disclosure, especially for inventions in AI, blockchain, and quantum systems. While the bar is high, the intent is to ensure enablement - a welcome move for serious innovators and a guardrail against overbroad or speculative claims.
Business Methods: The Line Remains Firm:
Despite advances in tech integration, the guidelines reiterate that business methods “per se” remain excluded under Section 3(k), regardless of any technical implementation.
Strengths and Gaps: A Balanced Take
The guidelines 2.0 mark a considerable improvement in both structure and substance. However, like any evolving regulatory instrument, it presents a mix of strengths and unresolved gaps that merit close attention.
KEY STRENGTHS:
1. Case Law-Backed Structure Enhances Trust
One of the most commendable aspects of the revised guidelines is the deliberate anchoring of examination standards in Indian judicial precedents. By explicitly referencing landmark decisions such as Ferid Allani v. Union of India and Ericsson v. Lava, the Patent Office has aligned its internal examination protocols with established jurisprudence.
This move strengthens legal certainty for applicants, reduces unpredictability in prosecution, and demonstrates that examination is not being conducted in a policy vacuum but within a legitimate judicial framework.
2. Use of Realistic, Sector-Specific Examples
Annexure I provides over 20 illustrative examples, including both patent-eligible and non-eligible inventions. These examples span:
AI-based image processing tools
Consensus mechanisms in blockchain
Non-patentable business rule automation methods
For years, the absence of contextually grounded examples has been a sticking point for practitioners. This annexure helps bridge the gap between abstract legal standards and real-world applications, especially for early-stage startups and academic inventors who may not be familiar with nuanced patent drafting.
3. Greater Clarity on AI and Cryptographic Disclosures
The guidelines now articulate a clearer expectation regarding technical disclosure for inventions involving:
Machine learning and deep learning models (e.g., neural network architecture, training datasets, loss functions),
Blockchain implementations (e.g., proof-of-work/proof-of-stake algorithms, cryptographic protocols),
Quantum algorithms and hardware interaction.
This raises the bar for claim enablement and discourages speculative filings that lack substantive technical content, thereby improving the overall quality of applications and strengthening enforceability of granted patents.
REMAINING GAPS AND AREAS OF CONCERN:
1. Blurring of Key Technical Terms
The draft continues to use the terms “technical effect”, “technical contribution”, and “further technical effect” without clearly defining their boundaries or relationships. While these terms are borrowed from European and Indian precedents, their overlapping use within the same document creates ambiguity.
For instance, is a “technical contribution” a subset of “technical effect” or an independent criterion? Should they be assessed cumulatively or sequentially? Without internal consistency, the door remains open to subjective interpretation.
2. Lack of Objective Standards for “Technical Effect”
The guidelines list acceptable “technical effects” (e.g., enhanced security, reduced latency, improved hardware performance), but they do not lay down any measurable benchmarks.
Unlike jurisdictions like the EPO that sometimes consider quantifiable improvements (e.g., X% increase in processing efficiency), the Indian draft leaves it to examiner discretion, which can result in inconsistent grant/rejection decisions across different technology centers or even within the same art unit.
This gap is especially problematic in AI, where improvements are probabilistic or stochastic in nature and difficult to express deterministically.
3. High Risk of Examiner Subjectivity
Because the final determination of whether an invention has a “technical effect” or is merely a software implementation depends heavily on examiner judgment, the lack of rigid metrics or checklists can lead to:
Varied examination outcomes for similar claims,
Increased burden on applicants to file extensive responses or appeals,
Reduced confidence in the process, especially for first-time filers or international applicants unfamiliar with Indian practice.
To mitigate this, the guidelines could benefit from an appendix that lists checkable criteria or decision trees, providing examiners with more consistent tools for analysis.
What Should Innovators Do?
While the Draft CRI Guidelines 2.0 are a leap forward, the long-term success of their implementation will hinge on refining definitions, clarifying conceptual overlaps, and providing both quantitative and procedural guidance to reduce examiner variability. Stakeholders should view this as an opportunity to contribute to a more predictable and innovation-aligned IP environment in India.
Feedback Deadline: July 7, 2025
Subject Line: “Comments on Revised Draft CRI Guidelines 2025 - Version 2.0”
Conclusion: A Welcome Update, with Work Left to Do
The June 2025 update to the CRI Guidelines is a commendable and forward-thinking document. It shows the Indian Patent Office’s responsiveness to community feedback and its intent to keep pace with disruptive technologies. The hope now is that the feedback submitted will help convert this draft into a practical, innovation-forward instrument that strikes the right balance between patent exclusivity and public clarity.
Author: Divyendu Verma
----------------------------------------------------------------------------------------------------------------------------------
Footnotes:
1. Comments on Revised CRI Guidelines 2025 can be submitted until July 7, 2025.
2. Reference case laws include Ferid Allani, Microsoft Technology Licensing, Ericsson v. Lava, and others.
3. Seven Stambhas refer to the claim structure analysis method recognized by Delhi High Court in Ericsson vs Lava.
4. For more details, visit: https://ipindia.gov.in/Home/NewsDetail/42
Kommentare