This issue of the Bulletin covers two crucial developments from the month of April 2026 – (1) The rejection of the DABUS patent application by the Indian Patent Office; and (2) The notification of the Online Gaming Rules 2026.

Indian Patent Office Rejects DABUS Patent Application

In a decision that follows similar precedents from jurisdictions such as USA, UK, Australia, Germany and Switzerland, the Indian Patent Office refused a patent application which sought to formally name an AI system – DABUS as a sole “inventor.” The application by Dr. Stephen Thaler sought to credit an AI LLM – DABUS (Device for the Autonomous Bootstrapping of Unified Sentience) for the autonomous invention of a unique food container featuring fractal profiles and a flashing beacon designed to attract attention in emergencies.

Unlike many patent offices worldwide that rejected the DABUS applications strictly on procedural grounds regarding the lack of an “inventive step,” the Indian Patent Office conducted a thorough examination of both the technical merits of the claims and the definition of an “invention” under §2(1)(j) r/w §2(1)(ja) of the Patents Act, 1970. The refusal of the patent rests on several grounds.

• First, the statutory construction of “inventor” under the Patents Act, 1970. The Office applied Sections 6, 7, and 10 of the Act directly to the facts. These provisions of the Act presuppose that an inventor is capable of holding legal rights, assigning them and being represented, which is why Section 6 only permits applications filed by the true and first inventor, their assignee, or a legal representative. Accordingly, the Office held that since a machine cannot execute a declaration, or possess legal rights, DABUS application failed to meet the statutory requirements.
• Second, the rejection of the ownership argument. The applicant contended that ownership of DABUS of its source code, hardware and outputs conferred entitlement to the invention, similar to an employer claiming rights over an employee’s work. The Patent Office rejected this analogy citing Section 7(2), which requires proof of right derived from a legally valid inventor. Mere ownership of a machine does not satisfy this requirement.

On a similar line of reasoning, the applicant also sought to rely on a declaration under PCT Rule 4.17(ii) but this too was rejected on the ground that a procedural declaration in the international law cannot override the mandatory substantive requirements u/s 6, 7 and 10 of the Patents Act, 1970 under Indian domestic law.

• Third, the absence of an “inventive step” originating from a human mind as mentioned under Section 2(1) read with Section 2(1)(ja). The Assistant Controller held that “since inventive step is considered from the point of view of a POSITA (Person of Ordinary Skill in the Art), the conception of invention should also originate from a human mind.” This reasoning essentially grafts a human cognition requirement onto the inventive step analysis by arguing that the benchmark for evaluating non-obviousness (the hypothetical skilled human) presupposes that the act of invention itself was performed by a human.
• The Office further rejected the argument that the absence of an express statutory bar allows non-human inventorship, holding that patent rights are statutory rights and inventorship must be recognised only within the framework enacted by Parliament. Legislative reform proposals, however well-reasoned, could not alter the present legal position.

There is a growing concern that this approach may create a legal vacuum or an “inventorship gap” wherein even the human programmer cannot truthfully claim to have conceived the specific output, especially given the lack of a global or national consensus on the threshold for human involvement in AI-generated works. For instance, Daria Kim, in her analysis of the German DABUS decision, highlights a glaring contradiction – the German court allowed Thaler to be named the inventor to preserve the patent’s economic value (note that this was not done in India), even though Thaler explicitly testified that he had no influence over the invention’s genesis! By allowing a human to claim an invention they admit they did not conceive, courts and patent offices are implicitly prioritizing the economic incentivization of patents over the genuine factual genesis of the invention.

One of the major goals of patent law is to maximize public benefit by incentivizing the disclosure of new technologies. In light of this, if valuable AI-generated breakthroughs in sectors such as new pharmaceutical drugs or advanced materials cannot be patented, corporations will inevitably resort to hoarding these discoveries as trade secrets. To take an illustrative example, in the pharmaceutical market, while the final AI-generated product (the drug) must be exhaustively disclosed to regulatory bodies (like India’s CDSCO) to prove safety and efficacy, the processes, training data, and methods used to discover, formulate, and manufacture that drug using LLMs may definitely be protected as ‘trade secrets.’ This may end up stifling public disclosure of beneficial technologies, hinder follow-on research and thereby slow global technological progress. However, this concern remains entirely speculative and it remains to be seen how the application of AI impacts innovation given the aforementioned challenges.

Ultimately, the Indian Patent Office’s refusal of the DABUS application reaffirms the established position that under current statutes, the title of “inventor” remains limited to human beings. However, with the rapid development of Generative AI, the burden (¶8.2-¶8.7) seems to be piling up for the Indian Parliament to introduce provisions specifically addressing AI-related inventions.

Notification of the Promotion and Regulation of Online Gaming Rules, 2026

The Promotion and Regulation of Online Gaming Rules, 2026 were notified in April 2026 and came into force on 1 May 2026. They operationalise the new gaming framework by creating a structured regime for registration, oversight, user protection, and grievance handling, while sharply separating e-sports and online social games from online money games. The practical effect is to move Indian regulation from ad hoc state control to a central, rule-based framework.

Their implication is a decisive policy shift. The Rules move beyond mere regulation of online money gaming, sitting within a statutory architecture that treats stake-based online play as a distinct regulatory harm, rather than as a protected commercial activity. That is a significant deviation because the debate around legality of money games was situated around the skill-chance distinction. Under the new regime, the decisive question is no longer whether a game involves skill in some measure, but whether it falls within the prohibited online money category. In State of Bombay v RMD Chamarbaugwala, the Supreme Court treated gambling as outside constitutional protection and developed the basic distinction between gambling and competitions involving substantial skill, with the latter being protected under Article 19 of the Constitution. In K Satyanarayana, the Court held that rummy, a game often played in Real Money Games, is a game predominantly of skill. The online gaming framework sidesteps this jurisprudence by building a separate statutory regime around real-money staking and regulatory control. Concerns along these lines have already been raised by commentators at the time of the enactment of Promotion and Regulation of Online Gaming Act in 2025, arguing that the blanket prohibition is constitutionally suspect and economically overbroad.

Although the Draft Rules were released for public consultation in October 2025 and elicited extensive stakeholder criticism, no consolidated report or reasoned account of how such feedback was incorporated accompanied the final notification.

In international terms, India is moving toward a model that is more categorical in its treatment of money gaming. In the UK, online gambling businesses must verify a player’s age and identity before gambling. Germany’s State Treaty on Gambling permits online betting and online poker under a licensing system. Switzerland allows online gambling only through licensed casinos, with approval for each game offered online. The common thread is tight licensing, identity checks, and consumer protection instead of a blanket ban. India’s framework goes further by drawing a harder line around online money gaming itself.