The deployment of generative artificial intelligence in industrial design has moved from experimental use to a core business function. For R and D teams and founders, the ability to generate thousands of product iterations in seconds offers a clear commercial advantage. However, this technical capability faces a significant legal bottleneck at the Indian Patent Office. The core question is whether a design created by an algorithm qualifies for statutory protection under the Designs Act, 2000.
Industrial design protection in India is built on the premise of human creativity applied to an article. As AI systems become more autonomous, legal friction increases. This article examines the statutory framework, examiner practice, and filing strategy for AI influenced designs in 2025 and 2026.
The Intersection of Generative AI and the Designs Act, 2000
Statutory Definition of Design and the Aesthetic Threshold
Section 2(d) of the Designs Act, 2000 defines a design as the features of shape, configuration, pattern, ornament, or composition of lines or colours applied to any article by any industrial process, which in the finished article appeal to and are judged solely by the eye.
Two requirements are central.
First, the subject matter must be visual.
Second, the feature must not be dictated solely by function.
AI systems are highly capable of generating visually appealing forms. However, the Act excludes any mode or principle of construction. If a generative design is primarily the result of structural or performance optimisation, the resulting shape may be treated as functional rather than aesthetic.
Based on current judicial interpretation, including the Supreme Court’s 2025 ruling in Cryogas Equipment v. Inox India, where the dominant purpose of a feature is functional utility rather than visual appeal, design registration is barred.
The Author Conundrum under Section 2(g) and Section 2(j)
Section 2(g) defines original as originating from the author of the design. Section 2(j) defines the proprietor as the author or the person for whom the design is executed.
The Designs Act does not define author as a natural person. However, Indian jurisprudence has consistently assumed human authorship.
Indian copyright practice, including the refusal to recognise RAGHAV AI as an author, indicates a strong policy position that authorship must vest in a human.
If an AI is the sole creator, there is no legally recognised author. If there is no author, there can be no proprietor. This creates a foundational vulnerability in purely AI generated designs.
Navigating the Novelty and Originality Mandate
Absolute Novelty in the Era of Algorithmic Proliferation
India follows an absolute novelty standard. Any prior publication anywhere in the world destroys novelty.
AI generated images are frequently uploaded to public platforms during experimentation. If any version of a design is disclosed before filing, registration will fail under Section 4(b).
The scale of AI content online creates two risks.
First, accidental self publication by internal teams.
Second, hidden prior art that may surface during cancellation proceedings.
Strategic practice requires strict internal confidentiality controls over AI outputs selected for filing.
Distinguishing New from Algorithmic Variation
Section 4(c) bars designs that are not significantly distinguishable from known designs.
Generative models often interpolate between existing designs. This creates outputs that are sophisticated trade variations rather than genuinely new visual expressions.
Examiners increasingly object where an AI generated form appears to be a predictable variation of industry standard geometry.
Applicants must demonstrate a clear visual departure from known design language.
Strategic Analysis: AI-Generated versus AI-Assisted Designs
The Human in the Loop Requirement for Registrability
Based on current IPO examination practice, a critical distinction is drawn between:
· AI generated designs where the system acts autonomously
· AI assisted designs where a human directs, selects, and refines the output
Where a human performs:
· prompt formulation
· selection from multiple outputs
· manual modification using CAD tools
the design is treated as originating from the human.
This satisfies Section 2(g.
Purely autonomous generation without meaningful human control creates a high risk of rejection.
Evidentiary Documentation and Design Logs
For AI assisted designs, contemporaneous documentation is essential.
Design Logbook Evidence
· Prompt histories
· Rejected iteration records
· Selection rationale
· Manual modification files
· Version control archives
This documentation serves as evidence of human authorship in both prosecution and enforcement.
Procedural Reality at the Indian Patent Office
Form 1 Disclosures and Applicant Declaration
Form 1 requires identification of the designer.
Listing an AI system as designer will result in rejection or abandonment.
Current practice is to list the supervising human designer.
There is no mandatory requirement to disclose AI usage. However, if questioned, the applicant must demonstrate human authorship.
Anticipating Examiner Objections on Non Human Authorship
Examiners increasingly identify AI generated aesthetics.
If originality is questioned, responses must establish:
· human intellectual contribution
· creative judgment
· tool based use of AI
The AI must be characterised as an instrument, not as the originator.
Comparative Jurisprudence: India, USPTO, and EPO
The Thaler Precedents and Global Consensus
Across jurisdictions:
· USPTO requires human inventorship
· EPO requires human inventorship
· UK Supreme Court requires human inventorship
India is aligned with this consensus.
Without legislative amendment, AI cannot be recognised as a designer.
Convergence toward Human Centric Design Protection
The emerging standard is significant human contribution.
Design protection follows control, not computation.
Risk Management and Enforcement for AI Influenced Designs
Validity Challenges under Section 19
Grounds include:
· prior publication
· lack of novelty
· lack of originality
Autonomous AI generation increases vulnerability on all three grounds.
The Eye of the Consumer Test in Infringement
In infringement suits, the only test is visual similarity.
The use of AI by either party is irrelevant.
Liability turns solely on visual impression.
Checklist: Protecting AI Driven Industrial Designs in India
Pre Filing Risk Mitigation Framework
|
Phase |
Action |
Purpose |
|
Audit |
Identify AI contribution level |
Section 2(g) risk |
|
Documentation |
Preserve logs and versions |
Prove authorship |
|
Confidentiality |
Avoid public disclosure |
Preserve novelty |
|
Filing |
Name human designer |
Procedural compliance |
|
Refinement |
Manual edits to AI output |
Distinguish from prior art |
|
Search |
Image novelty search |
Section 19 defence |
Frequently asked questions (FAQs)
Can an AI be named as a designer in India
No. Only natural persons can be designers.
Are AI generated designs registrable
Yes, if human authorship is established.
Will IPO reject designs that look algorithmic
Not automatically, but originality scrutiny increases.
Does AI use affect term of protection
No. Term remains ten plus five years.
Is prompt engineering sufficient for authorship
Only if coupled with selection and modification.
Can competitors challenge AI designs
Yes, on authorship and originality grounds.
Are GUIs generated by AI registrable
Yes, subject to the same rules.
Should AI usage be disclosed
Not mandatory, but internal proof is critical.
Is functional generative design registrable
No, if dictated solely by function.
Will Indian law change soon
No amendment as of 2026.