💥Mains Ready By December. Smash Mains & Smash PYQ Admissions Open

Subject: IPR

  • Can AI claim copyright for original work? A question of authorship

    Can AI claim copyright for original work? A question of authorship

    Why in the News

    India’s Copyright Office has rejected an application seeking copyright registration for an artwork generated by an artificial intelligence (AI) system. The application was filed by American computer scientist Stephen Thaler for a work titled ‘A Recent Entrance to Paradise’, which he said had been generated autonomously by his AI system DABUS. The application named DABUS as the author and Thaler as the owner of the copyright. The order is among the first Indian decisions to address who, if anyone, is the author when an AI system generates a work. The tension it exposes is that the Office found the image original enough to qualify for protection while holding that the entity that produced it cannot be an author.

    What is DABUS?

    1. The system: DABUS stands for Device for the Autonomous Bootstrapping of Unified Sentience, an AI system developed by Thaler.
    2. The claim made for it: The application asserted that DABUS had generated the artwork autonomously, rather than as an output directed by a human operator.

    What did the application claim and what did the Office ask?

    1. The filing: Thaler applied in 2022 to register copyright in the artwork.
    2. The first question put to him: The Copyright Office asked whether an AI system could legally be recognised as an author under the Copyright Act, 1957.
    3. The second question: It also asked who should be treated as the author if the work was indeed generated using AI.
    4. The offer he refused: During the proceedings the Office allowed Thaler to amend the application and identify himself as the author. He declined, and continued to insist that DABUS be recognised instead.

    How does the Copyright Act, 1957 treat originality?

    1. The three separate questions: The Act answers whether a work is original, who its author is, and who owns the copyright, and these are distinct questions rather than one.
    2. The protection provision: Section 13 protects original literary, dramatic, musical and artistic works.
    3. The Act does not define originality: The Copyright Office therefore interprets it from Eastern Book Company v. D.B. Modak.
    4. The judicial test: The Supreme Court in that case held that a work need not be novel or groundbreaking to receive copyright protection. It must show at least a minimum degree of creativity, and it cannot be merely copied or mechanically reproduced.

    How does the Act treat authorship and ownership?

    1. The authorship provision: Section 2(d)(vi) identifies the author of a computer generated work as “the person who causes the work to be created”.
    2. The disputed phrase: The dispute was over whether that phrase refers to the machine producing the output or to the person creating and operating the system.
    3. First ownership: Section 17 states that the author is generally the first owner of the copyright.
    4. Transfer: Sections 18 and 19 allow copyright to be assigned or transferred through legally recognised agreements.
    5. What the structure assumes: The Office noted that these provisions are built around legal persons who can hold rights, transfer them and enforce them.

    What did the Copyright Office decide?

    1. Originality was satisfied: The Office found that the image generated by the AI was original enough to qualify for copyright protection.
    2. Authorship is a legal status: The Act treats authorship as a legal status carrying rights and responsibilities, and an AI system, however sophisticated, does not presently possess such recognition under Indian law.
    3. The tool test: To interpret who “causes” a computer generated work to be created, the Office looked to American copyright cases distinguishing between a tool and the person handling it.
    4. DABUS as the tool: Although DABUS generated the final image, it did so within a system designed and set in motion by Thaler, so DABUS was treated as the tool and Thaler as the person who legally caused the work to be created.
    5. Person means natural or juristic: Where an Act refers to a “person” it usually means a natural person or a juristic person such as a company, an entity capable of owning property and entering contracts. DABUS is not a recognised juristic person.
    6. The outcome: Thaler was held to be the person capable of being identified as the statutory author, so the application as filed did not meet the criteria under the Act.

    Why was the fallback request also rejected?

    1. What was sought: Thaler asked in the alternative that DABUS be recorded as the technological generator of the work.
    2. The register cannot confer status: The Office held that the register could not be used to indirectly confer legal status on an AI system.
    3. A procedural ground as well: No proper application seeking such an entry had been made.

    What has the order left open?

    1. A future application can succeed: The order leaves open the possibility of a fresh application that identifies the author in the manner the Copyright Act, 1957 requires.
    2. The change of law is reserved: Any broader change in the law would have to come from Parliament.
    3. The stated limit on administrative power: The order records that whether legal personhood or authorship should ever be extended to autonomous artificial intelligence “remains a policy decision strictly reserved for Parliament, and cannot be introduced via administrative reinterpretation”.

    Challenges to fitting AI generated works into copyright law

    1. Human contribution is not measurable at the point of registration: A registrar cannot tell from the output whether a prompt involved creative choice or a single instruction. Eg. The United States Copyright Office refused registration for the AI generated images in the comic ‘Zarya of the Dawn’ while protecting the human written text and arrangement.
      The Fix: Require a disclosure of AI involvement and of the specific human contribution as a mandatory field in the registration application.
    2. Training data use is unresolved: Models are trained on protected works without licence, so the lawfulness of the input sits behind every question about the output. Eg. Indian news publishers and a music industry body have sought to intervene in the Delhi High Court proceedings against OpenAI on this ground.
      The Fix: Legislate a statutory text and data mining exception with a transparency obligation on training corpora, so the boundary is set rather than litigated case by case.
    3. Ownership defaults to the operator rather than the investor: Treating the person who causes creation as the author leaves the platform, the model developer and the user with competing claims over the same output. Eg. Generative service terms typically assign output rights to the user by contract, which no statute confirms.
      The Fix: Make the allocation of rights in computer generated output a default statutory rule that contracts may vary, rather than leaving it to terms of service alone.
    4. Term of protection has no anchor without a human author: Copyright duration runs from the author’s lifetime, which cannot be computed where the generating entity does not die. Eg. The United Kingdom sets a fixed 50 year term for computer generated works precisely to avoid this problem.
      The Fix: Provide a fixed term measured from the date of creation for works with no identifiable human author.
    5. Enforcement needs an accountable person: Liability for infringing output, and standing to sue over it, both require someone the law can reach. Eg. An autonomously generated image that reproduces a protected character leaves no party with a stated duty under the current provision.
      The Fix: Attach statutory responsibility for infringing output to the person who deployed the system, mirroring the authorship rule the Office has applied.

    Conclusion

    The order settles who the author is and leaves untouched what the author did. A work the law accepts as original was produced by a process its named author did not perform, and the statute has no category for that gap. Parliament is the only body that can create one. The point to watch is whether computer generated works are taken up as a legislative question, or whether the issue keeps returning through individual registration applications and appeals against their refusal.

    Back2Basics

    1. Enactment: The Copyright Act, 1957 came into force in January 1958 and is India’s governing copyright statute.
    2. Administration: It is administered through the Copyright Office, which functions under the Department for Promotion of Industry and Internal Trade.
    3. Coverage: It protects literary, dramatic, musical and artistic works, along with cinematograph films and sound recordings.
    4. Registration is optional: Copyright arises on creation of the work, and registration serves as evidence rather than as the source of the right.

    [2014, GS3, 12 marks] In a globalised world, intellectual property rights assume significance and are a source of litigation. Broadly distinguish between the terms – copyrights, patents and trade secrets.”

  • How is the government of India protecting traditional knowledge of medicine from patenting by pharmaceutical companies?

    India’s traditional medicinal knowledge includes thousands of formulations and approximately 45,000 plant species, but faces biopiracy threats from multinational companies patenting indigenous resources without consent or compensation.

    Government Initiatives to Protect Traditional Knowledge

    Traditional Knowledge Digital Library (TKDL):

    Translates ancient medicinal texts from Sanskrit, Urdu, Tamil, Persian and other languages into English, French, German, Spanish, and Japanese for global patent examiners.

    Contains over 4.48 lakh formulations, including Ayurveda, Unani, Siddha, Sowa Rigpa, and Yoga knowledge systems.

    CSIR-TKDL actively files pre-grant oppositions and third-party observations; 283 patent applications were refused, amended, or withdrawn using TKDL evidence.

    The Biological Diversity Act, 2002: Mandates that any foreign individual or commercial entity seeking to use India’s biological resources or traditional knowledge must obtain prior approval from NBA.

    National Biodiversity Authority: NBA is a statutory body implementing the Biological Diversity Act, 2002 to protect India’s biological resources and traditional knowledge.

    People’s Biodiversity Register (PBR): Administered by the NBA, PBR serves as a formal tool for recording and maintaining comprehensive localized data on biological resources and their medicinal uses.

    Access and Benefit Sharing (ABS) agreements:

    Companies using Indian bio-resources must share royalties or benefits with the National Biodiversity Authority.

    These funds support local Biodiversity Management Committees and tribal communities.

    The Patents Act, 1970:

    States that an invention which is traditional knowledge, or an aggregation or duplication of known properties of traditionally known components, is not patentable.

    Mandates disclosure of the source and geographical origin of biological materials used in patents, with details shared with the NBA.

    Protection of Plant Varieties and Farmers’ Rights (PPV&FR) Act, 2001: Protects the rights of local communities and farmers over their traditional crop and medicinal plant varieties.

    By safeguarding indigenous medical heritage through the NBA and TKDL, India directly advances SDG 3 (Good Health and Well-being) and SDG 15 (Life on Land) while protecting local community rights.

  • What is the present world scenario of intellectual property rights with respect to life materials? Although, India is second in the world to file patents, still only a few have been commercialized. Explain the reasons behind this less commercialization.

    IPR grants legal rights over innovations, while life materials include genes, microorganisms, and GMOs. Their intersection determines ownership and commercialization of biological resources, shaping biotechnology, healthcare, agriculture, and innovation-driven economic growth.

    Present World Scenario of IPRs with Respect to Life Materials

    Biotechnology: Increased patents on GMOs and gene-editing technologies, though patent laws differ across countries. Eg- CRISPR-Cas9 patents in the US and restrictions in the EU.

    Ethical concerns: Patenting genes and life forms can create monopolies and limit public access to healthcare and seeds. Eg- Myriad Genetics BRCA1 gene patent case.

    Developing nations’ approach: often oppose patents on essential medicines and biological resources. Eg- India rejected Novartis Glivec patent under Section 3(d).

    TRIPS and global standards:

    The TRIPS Agreement requires patent protection but allows safeguards for public health and biodiversity.

    With the WTO moratorium ending after MC 14 Meet, countries can now challenge public-health measures like compulsory licensing for harming expected profits.

    Open-source movements: Open-access biological initiatives encourage collaborative innovation and protect farmers’ rights. Eg- Open Source Seed Initiative.

    Biopiracy: Unauthorized patenting of biological resources and traditional knowledge exploits indigenous communities without fair compensation.

    Reasons for Low Commercialization in India

    Weak industry-academia linkage: Limited collaboration between research institutions and industries restricts market adoption. Eg- About 13.8% of CSIR patents are licensed.

    “Valley of Death” funding gap: Indian universities lack sufficient funding to scale laboratory research and prototypes into commercially viable products through testing and trials.

    Weak Patent Quality: Many patents suffer from vague claims, weak disclosures, or insufficient novelty, making them vulnerable to litigation and revocation.

    Slow regulatory machinery: Patent approvals and clearances in India often take 5-7 years, delaying commercialization and reducing technological relevance.

    Complex tech-transfer policies: Fragmented institutional IP policies create legal uncertainty, discouraging industry partnerships.

    Lack of Skilled IP Management: Limited expertise in licensing, prior-art research, and market-oriented commercialization, causing many patents to remain commercially unused.

    Misaligned objectives: Universities and researchers prioritize patent filings for rankings and grants, while industries seek scalable, market-ready technologies.

    Low absorptive capacity: Most universities lack strong innovation ecosystems and technology-transfer infrastructure beyond elite institutions like IITs.

    Poor commercialization infrastructure: India lacks strong incubators and technology-transfer systems.

    Global competition: Indian innovations face competition from dominant multinational corporations. Eg- Pfizer global market dominance.

    Inadequate Innovation Ecosystem: Support systems such as advanced laboratories, industry mentors, commercialization hubs, and global market integration remain uneven across regions.

    Way Forward

    Shift from quantity-driven patenting to quality-driven innovation by rewarding commercially viable and genuinely novel research.

    Strengthen industry-academia collaboration through technology transfer offices, IP centres, and startup incubation ecosystems. E.g Bayh-Dole model of the United States.

    Emulate China’s metrics-based databases, using big data analytics to isolate high-value patents

    Develop specialized biotechnology and pharmaceutical IP commercialization hubs on the lines of innovation clusters in South Korea and Israel.

    Utilize the 2024 Patent Rules, advance renewal discounts, and expanded startup facilitator schemes to protect emerging technologies.

    Align academic incentives away from mere patent counts toward innovation impact, technology transfer, and market adoption.

    Enhance venture capital support, FDI confidence, and startup financing by ensuring strong and enforceable intellectual property rights.

    Promote uniform state-level IP policies, single-window commercialization portals, and support for SMEs and rural innovators.

    With the above measures India can convert its patents into drivers of innovation, technological self-reliance, and the vision of Viksit Bharat 2047.

    Nuclear energy