Legal Reasoning Questions for CLAT | QB Set 50

We have had art for far longer than we have had copyright. The Statute of Anne, widely seen as the first copyright law, was passed in Britain in 1710, after the era of Shakespeare and Milton. The British brought copyright law to India in 1847. The current Copyright Act is from 1957. In 1710, the law granted authors a limited monopoly of 14 years, with the possibility of one renewal. The monopoly right would only vest if it was specifically registered and multiple copies of the book deposited for distribution among libraries and universities.

Under the current law, the monopoly right goes well beyond the act of publishing, vests automatically the moment ‘a work’ is created, and lasts for the author’s entire lifetime plus 70 years posthumously. So, the thousands of random Instagram posts and notebook doodles that you have made are all protected under copyright law for centuries. While the public domain was once the default, now a nearly-perpetual copyright monopoly is the default, regardless of the commercial potential of the work or ambitions of the creator. This fundamental change in the nature of the law has deleterious consequences.

Findings from a study

As part of a research project by LIRNEasia, a Sri Lankan think tank, we studied the data governance regimes of seven countries in South and Southeast Asia. With respect to copyright, we found that in four out of seven, the law made web search engines and AI training illegal. Web search engines need to copy as much of the Web as they can (a process called ‘crawling’), effectively creating a mirror copy of all that is reachable on the web through links, but permissionless copying is prohibited by copyright law. Except in the Philippines and Sri Lanka (which have a flexible ‘fair use’ exception) and India (which, in 2012, introduced a specific exception for ‘the transient or incidental storage’ for ‘providing electronic links, access or integration’), no other country in our study provided an exception, meaning AI training is effectively illegal in most countries we surveyed.

This does not make sense. Web search engines and AI models do not view copyrighted materials as scribblings or poems or art the way humans do; for programs, it is merely ‘data’ for statistical purposes. Recognising this, many jurisdictions such as the European Union, Japan and Singapore have adopted ‘text and data mining’ exceptions in their copyright regimes, while others such as Hong Kong and South Korea are in the process of doing so. Japan’s law allows for an exemption for “Exploitations not for enjoying the ideas or emotions expressed in a work” (i.e., use by machines), and permits “using the work in data analysis”. This is sensible: copyright was never meant to cover mechanistic uses.

By not allowing for a broad text and data-mining exception, India has created a pall of legal uncertainty over the collection of the training data for many forms of AI. And by not having a flexible, general and open-ended exception (as countries such as Singapore and the United States do), India ensures that copyright law will always hamper technological developments.

There are separate concerns around the outputs of generative AI substituting creative labour. But copyright is meant to be about encouraging creativity, not about protecting jobs. Further, copyright law has never prohibited learning from examples and imitating — every artist studies predecessors, and every writer reads widely. Technology has always displaced jobs — we have far fewer rickshaw pullers, telegraphists, pankhaawallahs, stenographers, lift operators, bank tellers, typesetters, darkroom technicians and draughtsmen — yet, it has created new jobs as well. The advent of photography reduced the demand for portraitists, but enabled new forms of creativity and access to knowledge. We do not know what the impact of generative AI will be: we might, in the future, need greater government grants for arts and culture, or to strengthen the cooperative movement, potentially funded by taxes from large AI companies. But these ought not be dealt with in copyright law.

Creativity, access should be promoted

What copyright law should protect, however, are contributions to the commons. Open-licensed AI models and datasets exemplify this — developers and researchers absorbing massive computational costs to create what enables others to be creative. These models add to the common heritage of mankind rather than subtracting from it. Copyright law should encourage such contributions, not hinder them with the same restrictions designed to prevent commercial exploitation. Governments are also uniquely positioned to curate high-quality locally-relevant datasets for public benefit, they should establish safe harbour provisions that protect such datasets from copyright claims, at least when used for training open-source models.

We have seen copyright law repeatedly being weaponised to block beneficial technologies under the guise of protecting creators. The Authors Guild in the U.S. used copyright to block Amazon Kindle’s “Read Aloud” function, despite it being assistive technology that enabled visually impaired persons to listen to books they had legally purchased. Current copyright law blocks technologies that could democratise access to knowledge, unleash creativity, and drive innovation — the very things that copyright was meant to foster. India’s hosting of the AI Summit is the moment to act: it must lead efforts for all nations to adopt flexible exceptions that serve creators and the public, rather than the copyright industry. We need to bring copyright law into the 21st century by returning to its roots.

(Source: The Hindu)

Question 1

The Copyright Act, 1957 provides automatic copyright protection the moment a work is created and extends protection for the lifetime of the author plus 70 years.

Raghav writes short poems on his private notebook and never publishes them. Ten years later, a tech company uses those poems (without permission) to train its AI language model by copying and analysing them statistically.

If India does not have a broad text and data-mining exception, which of the following is the most legally accurate position?

A. The company is not liable because the poems were never published.
B. The company is not liable because AI only uses the poems as data.
C. The company may be liable because copyright vests automatically upon creation and copying without exception is prohibited.
D. The company is liable only if the poems have commercial value.

Answer: C

Question 2

In a country where copyright law does not provide a “fair use” or “text and data mining” exception, a web search engine copies large portions of publicly accessible websites to create an index for users to search efficiently.

The website owners file a lawsuit claiming copyright infringement.

Based on the reasoning in the passage, which of the following best reflects the legal issue?

A. The search engine is automatically exempt because it improves public access to information.
B. The copying may be illegal because permissionless reproduction is prohibited under strict copyright law.
C. The search engine is protected because websites are in the public domain.
D. The lawsuit will fail because statistical use is never considered copying.

Answer: B

Question 3

Japan’s copyright law allows an exemption for “exploitations not for enjoying the ideas or emotions expressed in a work” and permits use in data analysis.

An AI company in Japan copies thousands of copyrighted novels solely to extract linguistic patterns for machine training, without reproducing expressive portions in outputs.

Which of the following is most consistent with Japan’s approach?

A. The company is liable because any copying is prohibited.
B. The company is liable because novels are creative works.
C. The company is not liable because the use is mechanistic and not for enjoying expressive content.
D. The company is not liable only if it shares profits with authors.

Answer: C

Question 4

A new generative AI tool produces artwork that resembles traditional portrait styles. A group of portrait artists argues that AI should be banned because it may reduce their employment opportunities.

Based on the reasoning in the passage, which principle is most relevant?

A. Copyright law exists primarily to protect employment in creative industries.
B. Copyright law exists to prevent technological change.
C. Copyright law is intended to encourage creativity, not to guarantee jobs.
D. Copyright law must prohibit all imitation of artistic styles.

Answer: C

Question 5

The government proposes creating a publicly funded, open-licensed dataset for training open-source AI models. It also plans to introduce a safe harbour provision protecting such datasets from copyright claims when used for public-benefit AI research.

Which of the following best aligns with the reasoning in the passage?

A. The proposal undermines copyright entirely and is inconsistent with its objectives.
B. The proposal promotes contributions to the commons and aligns copyright with innovation goals.
C. The proposal is illegal because government cannot curate data.
D. The proposal protects only commercial AI companies.

Answer: B