\n<\/p>\n
.<\/span><\/p>\n This piece considers * An early draft of this piece was originally published here as ‘Artificial Intelligence: The Making or Breaking of Copyright Law<\/i>‘ on April 8, 2025.<\/span><\/span>\u00a0<\/span><\/p>\n \u00a0<\/span><\/p>\n Contents<\/span> 1\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 <\/span>Assisted Creativity and Copyrightability. <\/span>2<\/span><\/span><\/p>\n 2\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 <\/span>Privileged Creators and Trade. <\/span>4<\/span><\/span><\/p>\n 3\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 <\/span>The Statutory Anticipation of AI. <\/span>5<\/span><\/span><\/p>\n 4\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 <\/span>Acknowledged Authorship. <\/span>8<\/span><\/span><\/p>\n 5\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 <\/span>The Data underlying Computational Creativity<\/span><\/p>\n 6\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 <\/span>Rethinking the Scope of Protection. <\/span>15<\/span><\/span><\/p>\n \u00a0<\/span><\/p>\n \u00a0<\/p>\n Two Hephaestus\u2019 Nonetheless, It Focussing Copyright Perhaps A An Closely The In Strangely, Or Although \u00a0<\/span><\/p>\n While In The As As This Further, It Unfortunately, Unfortunately, The The How It Coherent Even Another In The The Protected Section There It It One The One Simultaneously, Copyright When In Ensuring Finally, Looking Such Copyright The Thus, In Assuming As The While As In Thus, On The Further, The Other
<\/span>Draft: 14 MAY
\n2025<\/span><\/p>\n
\nthe interaction of Indian copyright law and artificial intelligence, explores
\nhow the law could be amended to make it more compatible with AI, and suggests
\nthat AI has simply highlighted age-old concerns which the law has given rise to
\nrather than created a series of novel challenges.<\/span><\/p>\n
\n\u2018female forms\u2019 which \u2018mov\u2019d, and breath\u2019d, in animated gold\u2019 supported
\nHephaestus, \u2018Sovereign of the fire\u2019 and god of craftsmen, as he approached
\nThetis, Achilles\u2019 goddess mother, who was to commission armour for her son.
\nThese female forms had been forged by Hephaestus and, <\/span>we are told<\/span><\/a>, had \u2018voice, and sense, and science\u2019
\nbestowed upon them by the gods.[1]<\/b><\/span><\/span><\/span><\/p>\n
\nhelpers are widely believed to have been androids, and, yet, the Iliad<\/i> describes the figures which adorn the
\nwarrior Achilles\u2019 shield, also forged by the god of craftsmen, in remarkably similar
\nterms. This leads one to wonder if Hephaestus\u2019 women of gold were, perhaps, not
\nandroids at all but dynamic figures on greaves, armour protecting tibiae, which
\nquite literally supported Hephaestus by holding him upright and facilitating
\nmobility. Such an interpretation does not seamlessly align with the helpers\u2019
\ngrasp of science but neither does it disturb the practice by which living
\nbeings, whether gods or humans, may labour assisted to fabricate works which
\nwould, were they made in our own time, likely be eligible to be protected by
\ncopyright and, possibly, other intellectual property laws.<\/span><\/p>\n
\ndespite assisted creativity being well established, there is considerable
\ncontemporary debate about which uses of technology are legitimate. Social media
\nplatforms often host heated discussions in arts and crafts groups about which
\nhandicrafts merit being described as having been handmade. Many knitters, for
\nexample, appear to treat knitting using a loom with derision; they refuse to
\nrecognise items made using looms as truly being handmade. From their comments,
\nit appears that they view even manual looms, unlike the knitting needles they
\nthemselves use, as machines rather than as tools.<\/span><\/p>\n
\nis not entirely clear where the dividing line between a <\/span>tool<\/span><\/a> and a machine<\/span><\/a> lies. The former is defined by the
\nCambridge Dictionary as \u2018a piece of equipment that you use with your hands to
\nmake or repair something\u2019[2]<\/b>\u00a0while the
\nlatter is defined as \u2018a piece of equipment with several moving parts that uses
\npower to do a particular type of work\u2019.[3]<\/b>\u00a0The definitions provide little clarity since even common pieces of equipment like
\nelectric drills and some sewing machines, such as those which are handheld and
\nmanual, straddle the line between tools and machines.<\/span><\/p>\n
\non aids which help bring tangible works into being, the difference between
\n\u2018tools\u2019 and \u2018machines\u2019 matters not least because it is far more likely, in our
\ntime, that items made by hand rather than by machines would be considered to be
\neligible to be protected by copyright. In the case of looms, an
\nintricately-patterned scarf woven on a handloom used to make cloth would not be
\nconsidered to be anything but handmade and, if its design were original,
\npotentially copyrightable. A scarf made on a knitting loom, however, could
\neasily be alleged to not have been handmade by self-proclaimed purists of the
\ncraft and may struggle to cross the threshold of copyrightability since such
\nitems tend to have designs which are fairly predictable.<\/span><\/p>\n
\nlaw leans towards protecting the products of human labour especially if, in
\ntheir fabrication or creation, human ingenuity has been applied. In doing so,
\nit has come to protect a mixture of works, tangible and intangible, depending
\non the circumstances: not just the content of books and other literary works
\nbut also computer programmes (deemed to be literary works by the law),[4]<\/b>\u00a0musical
\nand dramatic works, sound recordings, films, artistic works (including
\nphotographs, sculptures, and works of craftsmanship) and computer-generated
\nworks. It requires these works to be either original or non-infringing
\ndepending on their nature,[5]<\/b>\u00a0to have
\nbeen created independently of other pre-existing or contemporaneous works,[6]\u00a0<\/b>and to not be so banal as to have their treatment of a subject be unavoidable.<\/span>[7]<\/b><\/p>\n
\nbetraying its origins as a right printers lobbied for, manifested through the
\ngenesis of Great Britain\u2019s seminal 1709 Statute of Anne, the law pertaining to copyright
\ntreats trade in books as a given, allowing the content of literary works as
\nwell as other works which can be printed on to paper or like media, physical or
\nelectronic, to retain copyright even if copies of the works are mass-produced
\nand, in the case of works of art, are in two-dimensional form.[8]<\/b>\u00a0In
\ncontrast to this, copyright law is less enthusiastic about trade in tangible
\nworks which would ordinarily merit copyright protection (especially if they are
\nthree-dimensional and visually-appealing); it tends to protect the works
\nthemselves rather than content embedded in them, and to limit the extent to
\nwhich they can be protected by the 1957 Copyright Act, the primary statute
\nwhich governs the subject in India, upon their having been legally reproduced
\nmore than 50 times by an industrial process,[9]<\/b>\u00a0which, in practice, is when such works tend to enter commercial channels.<\/span><\/p>\n
\nless cynical outlook would perhaps explain the disparities in the treatment of
\nvarious kinds of works, and copyright law\u2019s choice of which mass-produced works
\nto shun, not with reference to the origins of copyright law but with reference
\nto the purpose of the grant of copyright protection. Considered through a
\nutilitarian lens, that copyright law tends to protect original, contemporary books
\ncontaining literary, dramatic, musical, and artistic works without reference to
\nhow many copies of them are produced is unsurprising; copyright is ostensibly
\nintended to protect the revenue streams of authors and those works which
\nbenefit society at large, particularly if that benefit is achieved through the
\nfacilitation of access to academic learning, a purpose books are ordinarily
\nintended to help fulfil.<\/span><\/p>\n
\nintroduction to the earliest Indian copyright law,[10]<\/b>\u00a0an 1847 statute, noted that it would be expedient to establish a copyright law
\nin those parts of India governed by the East India Company to encourage
\nlearning[11]<\/b>\u00a0while,
\nover a century and a half later, the overriding purpose of the 2012 amendment
\nto the Indian copyright statute, the 1957 Copyright Act in force in the
\nRepublic of India since 1958, was to protect the interests of musicians and
\nlyricists in the film and music industry.<\/span><\/p>\n
\nlinked to the broader societal benefits of protecting copyrightable works is,
\nof course, the need to protect the creators of those works: in that swoop, the
\nlaw protects authors (and, in recent decades, performers),[12]<\/b>\u00a0especially if they are natural persons, by helping them monetise their work
\nthrough the grant of exclusive rights in their creations and performances.<\/span><\/p>\n
\nlaw\u2019s close links with the potential for monetisation, demonstrated not least
\nby the ease with which copyrights can be licensed, bought and sold, may have resulted
\nin those creators, who can competently leverage markets and social networks not
\nto mention navigate legal labyrinths, being able to financially benefit from
\nhaving their works be protected by copyright far more easily than those who
\ncannot.<\/span><\/p>\n
\npractice, this has meant that, in making choices about whose works merit
\nprotection, the law, anchored in patriarchy and ableism despite attempts in
\nrecent decades to dislodge it from its moorings, has developed a tendency to
\nprotect those works historically associated with men for the most part, and to
\nprioritise the protection of works of individual genius. As a corollary, those
\ncreative items which have historically been overwhelmingly made by women are
\nalso those which are least readily protected by copyright law: quotidian
\ncookery is unlikely to be protected but the creations of (once often male)
\nchefs may be eligible to merit protection, for example, and mere gardens tend
\nto remain unprotected but vistas shaped under the supervision of (once ordinarily
\nmale) landscape architects can often be protected.<\/span><\/p>\n
\nitems made using looms and needles can far more easily be exceptions to the
\ngeneral rule that copyright has a limited role to play in the protection of so-called
\nwomen\u2019s arts. Needlework, brocades, lace, and tapestries all have the potential
\nto be considered works of art, perhaps because, although the media differ,
\ntheir visual impact can often be rendered using paints which men have
\ntraditionally favoured.<\/span><\/p>\n
\nperhaps there is an older tale to tell, which copyright law has subconsciously
\nimbibed: Pandora, handed over to the titan Epimetheus, was moulded by
\nHephaestus using clay and water. In Theogony<\/i>
\nand Works and Days<\/i>, <\/span>Hesiod tells us<\/span><\/a> that she was animated, and given
\ngifts by the gods; the goddess Athene taught her needlework and weaving.[13]<\/b>\u00a0It may
\nnot be mere coincidence that these are the two \u2018women\u2019s\u2019 crafts which copyright
\nlaw can most easily be called upon to protect.<\/span><\/p>\n
\nshe is commonly referred to as the first woman, at no point does Hesiod tell us
\nthat Pandora is a woman. Instead, she is described more as a prototype for
\nwomen than as a woman herself. Although she is clearly said to have been given
\nthe ability to speak, not once does she actually speak in Hesiod\u2019s telling.[14]\u00a0<\/b>And, so,
\nit may well be that these crafts, which women once predominantly engaged in,
\nwere, in times of myth, passed on to them not by a woman but by an android, a
\nphysical manifestation of artificial intelligence, resulting in their being
\ntreated as more than banal \u2018women\u2019s work\u2019, the products of mindful creativity
\n(or, at any rate, creativity derived from an android courtesy of the gods)
\nrather than mindless drudgery (undeserving of legal protection).<\/span><\/p>\n
\nit has never been unusual for human beings to create works with the assistance
\nof technology, until very recently, it has not been common for technology to
\ncreate works unassisted only upon being prompted to do so by humans. Some of
\nthe technology which allows for the fabrication of tangible works, such as that
\nused for 3D printing, has been relatively easily accessible by the public since
\nabout 2009 when an early patent for fused deposition modelling, a 3D printing
\ntechnique, expired.[15]<\/b>\u00a0Other
\ntechnologies, such as generative AI, used to create \u2018intangible\u2019 works such as
\ntext, audible music, and visual art which are recorded on or enjoyed through
\nphysical and digital media without necessarily being tangible themselves, have
\nonly exploded in popular consciousness since about 2023, in no small measure,
\noff the back of the release of a free version of ChatGPT based on GPT 3.5
\ntowards the end of 2022.<\/span><\/p>\n
\nthe early 1990s, Mark Weiser pointed out: \u201cThe most profound technologies are
\nthose that disappear. They weave themselves into the fabric of everyday life
\nuntil they are indistinguishable from it.\u201d<\/span>[16]<\/b>\u00a0This has certainly held
\ntrue in the case of AI, as artificial intelligence is commonly called: many
\nforms of AI, such as grammar- and spell-checks, are ubiquitous being all but
\nuniversally used; they are rarely thought of as being artificial intelligence
\nat all. The partial invisibility of AI on one hand contrasts with inescapable
\nvisibility on the other hand brought about by news of developments in the area
\nfrequently coming to the fore of public discussion. This makes it difficult not
\njust to define AI in any meaningful way for the purposes of the law which will
\nultimately likely have to settle on a single set of standards and policies to
\ngovern legitimate uses of the technologies involved without finely categorising
\nAI and finetuning the law for each category, an endeavour which would, in any
\ncase, likely fail given that scientific developments invariably to outpace
\nlegal developments.<\/span><\/p>\n
\nemergence of GenAI, as generative AI tends to be referred to, in public use and
\ndiscourse seems, at first glance, to have been anticipated by a provision of
\nthe 1957 Copyright Act introduced to the statute in 1994 which states that \u2018in
\nrelation to any literary, dramatic, musical or artistic work which is
\ncomputer-generated, the person who causes the work to be created\u2019 is its
\nauthor.[17]<\/b>\u00a0However,
\nit may not be easy to identify which works could be considered to have been
\ncomputer-generated not least because a definition of \u2018artificial intelligence\u2019
\nhas proved to be elusive, and policy documents (such as the Report on AI
\nGovernance Guidelines Development<\/i> released by the Indian Ministry of
\nElectronics and Information Technology in January 2025)[18]<\/b>\u00a0have demonstrated a tendency to forge ahead and discuss how AI should be
\ngoverned without delving into the issue of what constitutes AI in the first
\nplace.<\/span><\/p>\n
\nfar as authorship is concerned, it is not immediately apparent that the 2018
\nPractice and Procedure Manuals published on the website of the Indian Copyright
\nOffice are consistent with the mandate of the 1957 Copyright Act simply because
\nthe manuals are non-committal about who the authors of musical works may be
\n(with reference to whether they are natural or legal persons),[19]<\/b>\u00a0indicate
\nthat the authors of artistic works must always be human,[20]<\/b>\u00a0and assert that only human beings can be authors of literary works.<\/span>[21]<\/b><\/p>\n
\nsuch, the law does not clearly allow for the authorship of entirely
\ncomputer-generated literary, dramatic,[22]<\/b>\u00a0musical
\nor artistic works to be attributed to humans. Nonetheless, it would appear that
\nthe absence of categorical assertion, perhaps inadvertently, allows for the
\nauthorship of such computer-generated works to be attributed to companies; it
\nis not clear how the statutory position of attributing the authorship of computer-generated
\nliterary, dramatic, musical and artistic works to the persons who cause them to
\nbe created and the aversion to such persons being natural persons in practice could
\notherwise be reconciled.<\/span><\/p>\n
\nis in contrast to the treatment which both the 1957 Copyright Act and the 2018 copyright
\npractice manuals have accorded to works which are often composite in nature,
\nnamely: sound recordings and cinematograph films (referred to as \u2018composite
\nworks\u2019 in this piece); the statute states that their producers are their
\nauthors[23]<\/b>\u00a0while the
\npractice manuals allow \u2018producing companies\u2019 to be named as their authors.[24]<\/b>\u00a0Thus, in
\nthe case of composite works created without the support of artificial
\nintelligence, it appears that copyright law allows both natural and legal
\npersons (or, in the case of the latter, corporate entities, at any rate) to be
\nconsidered to be authors since there is no restrictive statutory provision
\nlimiting author-producers to either natural or legal persons. The statute itself
\nis silent on the subject of who the authors of computer-generated composite
\nworks may be but, despite the absence of explicit prescription, it could
\nperhaps be inferred that either human beings or legal persons could be
\nconsidered to be the authors of computer-generated composite works simply given
\nthe absence, too, of explicit proscription.<\/span><\/p>\n
\nthere is no guidance in the 1957 Copyright Act about what \u2018causing a work to be
\ncreated\u2019 by a computer means for the purposes of the statute. As a result, it
\nis also possible that statutory silence could be interpreted to mean that no
\nperson, human or legal, is the author of composite computer-generated works,
\nand that the software programmes which generate them are to be considered their
\nauthors; the statutory requirement that a person be the author of
\ncomputer-generated literary, dramatic, musical, and artistic works does not
\napply to computer-generated sound recordings and cinematograph films. If it were
\nthe case that software programmes could be considered to be the authors of
\ncomposite works, it is unlikely that works so created would be protected by
\ncopyright.<\/span><\/p>\n
\nis important to determine who the authors of potentially copyrightable works
\nare simply because authorship is often only one short step away from ownership:
\nthe author of a work is usually the first owner of copyright in the work.[25]<\/b>\u00a0Currently, where the statute is unclear, and computational creativity has
\nplayed a role in bringing copyrightable works into being, the authorship and
\nownership of those works tend to be determined by standard form contracts which
\ngovern the use of the relevant software.<\/span><\/p>\n
\nthis is not a foolproof approach to determining issues of either authorship or
\nownership: not only may such agreements ultimately fail the test of
\nconscionability should they be challenged (which would make their provisions
\nunenforceable), but also, given that, as a group, they have often been drafted
\nin a field rife with statutory uncertainties, they may simply fail to be
\nconsistent with what are ultimately judicially interpreted to be statutory
\nimperatives. There is therefore good reason to amend the law to clarify who the
\nauthors and owners of works created by or with the assistance of computational
\ncreativity would be, ensuring that technology companies which are the most
\npowerful players in the field do not simply override the rights of end-users
\nand appropriate their putative rights.<\/span><\/p>\n
\nhaving come into being as a somewhat mercenary right which tradesmen lobbied
\nfor, copyright law does not have a solid foundational understanding of what
\nauthorship entails. Instead, to this day, it relies on a rudimentary understanding
\nof who may be recognised as an author (in its quest to facilitate trade) which,
\nentirely unhelpfully, defines the authors of literary, and dramatic works as
\nthe authors of those works,[26]<\/b>\u00a0and veers
\naway from popular understandings of authorship primarily to accommodate
\ncommercial demands through its recognition of producers as the authors of what
\nwould once necessarily have been expensively-made composite works.<\/span>[27]<\/b><\/p>\n
\nlaw tends to treat all natural creators at par, equally but not necessarily
\nequitably, occasionally making allowances for them not with regard to their own
\ninnate abilities but with reference to the kinds of works they may create; the
\nso-called Bollywood amendments which were introduced to the 1957 Copyright Act
\nin 2012, for example, primarily attempted to ensure that composers and
\nlyricists working in the film and music industry, being the authors of
\nunderlying works in cinematograph films and sound recordings, would be treated
\nfairly.[28]<\/b>\u00a0The same
\nset of amendments also introduced provisions related to disability to the
\nIndian copyright statute[29]<\/b>\u00a0although the
\nfact that it focussed on treating people with disabilities as consumers of
\ncopyrighted works rather than as creators of copyrightable works is inescapable:
\nthe statute enhances the ability of people with disabilities to engage with existing
\ncopyrighted content by allowing works to be reformatted for their benefit but
\nit contains no particular provision to enhance the ability of people with
\ndisabilities to step into the role of legally-acknowledged authors perhaps by
\nlowering the threshold of copyrightability should they create works using
\nassistive technologies.<\/span><\/p>\n
\ntechnological assistance available to authors has come to take many forms most
\nof which defy heterogeneous categorisation: amongst them are a combination of
\nphysical tools at the disposal of authors, artificial intelligence which may be
\nused to support (and occasionally displace) natural and legal persons who are
\ncreators, and assistive technologies aimed at enhancing the ability of people with
\ndisabilities to engage with the world but which can also aid currently-abled
\npeople to do so.<\/span><\/p>\n
\na broader range of allowances to support creative endeavours could be framed remains
\nan open question, and, so far, the best efforts of the law to align itself with
\ntechnological advancements, and to ensure that all creators are treated fairly,
\nhave proved to be inadequate. Volition is more frequently recognised in the
\ncontext of infringement than of authorship with its likely having underlain
\nboth the understanding that the commission of secondary copyright infringement
\n(by allowing a place to be used to communicate copyrighted works to the public
\nfor profit) only gives rise to liability if the person committing such
\ninfringement is aware or at least has reasonable grounds to believe that the
\ncommunication would be infringing[30]<\/b>\u00a0as well as the development of exceptions to liability for copyright
\ninfringement enjoyed by intermediaries which merely provide neutral networks
\nthrough which infringing copies of works may have been transmitted.[31]<\/b>\u00a0Nonetheless, \u2018volition\u2019 may be a useful tool for developing a theory of
\nauthorship.<\/span><\/p>\n
\nhas been suggested that the intention of authors at the time of creating
\ncopyrightable works may play a role in determining the legal protection
\naccorded to their works.[32]<\/b>\u00a0However, while
\ndiscussing the relationship between copyright and designs law, the Delhi High
\nCourt noted: \u201cThe artist\u2019s intent at the time of creation of the artistic work
\nis indecipherable at the best of times. Artists are governed more often than
\nnot by their emotions and moods and whatever be the intention at the time of
\nthe creation of the artistic work cannot, in our view, determine the nature of
\nprotection available to the artistic work.\u201d[33]<\/b>\u00a0That being the case, it is unlikely that authorial intention or the conscious
\nvolition of creators could be used as the basis to validate legal authorship in
\nIndian jurisprudence.<\/span><\/p>\n
\nextrapolated volition, the machine counterpart of conscious human volition,
\ntoo, is, by extension, unlikely to be meaningful in terms of acknowledging
\nauthorship for the purpose of copyright law. This aligns with the law\u2019s
\nreluctance to recognise humans as the authors of computer-generated literary,
\ndramatic, musical and artistic works but it does nothing to present a solution
\nto the question of which factors should be considered to determine how authors
\nshould be recognised given exponential increases in the use of technology,
\nparticularly artificial intelligence, in creative processes.<\/span><\/p>\n
\nthough Indian law seems inimical to recognising conscious volition as the basis
\nfor acknowledged authorship, it may be possible to argue that functional
\nvolition (understood with reference to the underlying processes through which
\ncopyrightable works may be brought into being by creators) could legitimately
\nform the basis on which authorship is acknowledged by the law. Not only would
\nthis help align aspects of copyright law relating to authorship and
\ninfringement with each other by having a common thread run through each of them
\nbut it would also allow for copyright to be shaped to account for the
\ncapabilities of individual authors and the allowances which may need to be made
\nto ensure that they, and their use of technology, are considered fairly.<\/span><\/p>\n
\nalternative may be to consider authorship not in positive terms by attempting
\nto recognise it with reference to the use of technology but in negative terms
\nby describing when authorship would not be acknowledged by law on account of the
\ncircumstances in which content had been created; as <\/span>Elena Cooper
\nhas pointed out: \u201c\u2026the challenges to authorship may stem from the objectives of
\nparticular practices, informed by their context (e.g. online community resource
\nas opposed to art gallery display), rather that the use of digital technology per
\nse<\/i>.\u201d[34]<\/b>\u00a0Such an
\napproach could align with an emphasis on functional volition, and recognise the
\ndisparity between syntactic and semantic comprehension demonstrated by the Chinese
\nRoom Argument[35]<\/b>\u00a0(in which,
\nin essence, it is imagined that a person who does not know Chinese may render
\nthe false impression of understanding the language by \u2018conversing\u2019 using Chinese
\ncharacters chosen on the basis of detailed instructions in English).<\/span><\/p>\n
\nso re-considering authorship with reference to the circumstances of creation, technology
\nper se<\/i> could at least be partially dislodged as a mast upon which
\ncopyright law\u2019s extremely rudimentary understanding of authorship pivots. For
\nexample, text written as part of an open project could simply not be protected
\nby copyright at all instead of having to be freed through various \u2018copyleft\u2019
\nlicences. Thus, copyright law could be restructured to potentially remove from
\nthe scope of copyright protection the sundry, make doctrinal issues about what \u2018copyleft\u2019
\nstands for (other than being opposed to conventional copyright) irrelevant, and
\nharken to another time when content was protected upon registration. However,
\nunlike registration, considering the circumstances of creation instead of the
\ncompletion of a formal legal process would not disadvantage those creators unfamiliar
\nwith the legal system or creators with specific requirements or disabilities
\nwhich were alleviated through the use of assistive technologies.<\/span><\/p>\n
\nmanner in which people use assistive technologies to create copyrightable works
\nranges from using technologies to generate creative content in its entirety to accomplishing
\ndisparate (subsidiary) tasks embedded within broad creative endeavours to
\ncreate works (undoubtedly of their own making) with the support of technology. Assistive
\ntechnologies themselves include a variety of tools from tangible mechanical apparatus
\nat one end of the spectrum to intangible artificial intelligence (embedded in
\nhardware though it certainly is) at the other end of the spectrum. While in the
\ncase of tangible tools, the difference between \u2018tools\u2019 and \u2018machines\u2019 plays an
\nimportant in determining the copyrightability of works, in the case of intangible
\ntools such as AI systems, questions regarding the copyrightability and the use
\nof AI must not only consider that difference but also, bearing it in mind, ask how
\nAI interacts with pre-existing copyrighted content, go on to question what
\nprecise use AI is put to by its users, and finally determine if the output
\nwhich comes into being with the involvement of AI merits copyright protection.<\/span><\/p>\n
\nfirst of these questions requires an analysis of what data or content is fed to
\nAI to train it to act in specific ways, for what purpose the AI deals with the
\ncontent, and how it treats the content it is fed to formulate outputs.
\nStrangely enough, if one were to disregard the fact that the statute does not
\ndeal with minor infractions,[36]<\/b>\u00a0it is
\njust one provision of the 1957 Copyright Act which deals with all these aspects
\nof the issue: Section 52 which enumerates exceptions to copyright infringement.<\/span><\/p>\n
\nworks other than computer programmes, as Section 52(1)(a) of the 1957 Copyright
\nAct says, may be used without authorisation for private or personal purposes
\nincluding research, for the criticism or review of the works themselves or
\nother works, and for the reportage of current events, current affairs and
\npublic lectures. The provision does not differentiate between commercial and
\nnon-commercial use, and although its emphasis on private and personal uses in
\nits first subsection may lean towards such a demarcation, it is also possible
\nthat open artificial intelligence systems used by closed groups may satisfy the
\nrequirement that use be \u2018private\u2019 as opposed to being public to benefit from
\nthe exception carved out by the provision. Notably, the other two subsections
\nof Section 52(1)(a) of the 1957 Copyright Act dealing with review and reportage
\ndo not contain analogous restraints. All three subsections do, however, require
\nthat unauthorised use be fair to fall within the scope of the provision; Indian
\nstatutory law does not define \u2018fairness\u2019 in this context but an understanding
\nof fairness drawn from \u00a7 107, Title 17, USC, which explains \u2018fair use\u2019 in the
\ncontext of US copyright law, has been incorporated into the Indian corpus juris <\/i>through case law.<\/span><\/p>\n
\n52 of the 1957 Copyright Act carries on listing exceptions to copyright
\ninfringement from subsection (1)(a) all the way to subsection (1)(zc), finally
\nending with subsection (2) which extends the application of the exceptions
\ncarved out in Section 52(1) both to translations of literary, dramatic, and
\nmusical works and to adaptations of literary, dramatic, musical, and artistic
\nworks in the same way that they apply to the works themselves. The exceptions
\ncover a wide range of subjects and circumstances but their tenor remains
\nremarkably consistent in that they explain how the copyright in specific kinds
\nof works can be used without authorisation, and with limited regard to the
\nidentity of users. Their consistency may make them appear to be rigid but their
\ninterpretation in the real world has proven to be less than inflexible. Indian
\ncourts have, for example, recognised \u2018transformative use\u2019 as a valid defence to
\ncounter allegations of copyright infringement even though the term quite simply
\ndoes not appear in the 1957 Copyright Act, much less in Section\u00a052.<\/span>[37]<\/b><\/p>\n
\nis also the fact to consider that the exceptions to copyright infringement
\nappear to step straight from one work being used or copied to create another in
\na manner where proximity would, but for their existence, lead to the commission
\nof infringement. While such an arrangement functioned smoothly in a world that
\nhad not seen the proliferation of AI, it does not work as smoothly in an
\nAI-ridden world.\u00a0<\/p>\n
\nis not one step from having so-called training data as AI inputs to the
\nmanifestation of an AI output. The lack of proximity between input and output,
\nwhen it comes to AI, could be read in two ways: either that proximity (closely
\nlinked to the expressive use of one work in the creation of another) is
\nrequired for infringement to occur and, therefore, the lack of proximity means
\nthat no infringement has occurred, due to which the exceptions to copyright
\ninfringement in the 1957 Copyright Act tend to ignore situations in which there
\nis no proximity, or, alternatively, that choosing not to deal with situations
\nwhich lack proximity is a considered choice meaning that the lack of proximity
\nis no defence to allegations of infringement.<\/span><\/p>\n
\nis extremely unlikely that there could have been much thinking along these
\nlines when the 1957 Copyright Act was first drafted simply because the relevant
\ntechnology had barely begun to be developed, and, despite the statute having
\nbeen amended several times since then, there does not appear to have been much,
\nif any, attention paid to the possibility of AI gorging content as training
\ndata. This is far from the first time that the law has had to find ways to
\ncatch up with technology, and although the initial steps to do so may be taken
\nby the courts, ultimately, the question of how to handle AI may well fall to
\nParliament.<\/span><\/p>\n
\nway in which the law could handle copyrighted content being used as training
\ndata would be to craft a new exception to copyright infringement to allow for
\nsuch use. A broad exception would, however, be unlikely to be fair simply
\nbecause it would facilitate extremely well-funded AI companies using content
\nbelonging to others without either authorisation or remuneration. If the choice
\nto expand the statutory exceptions to copyright infringement were made, it
\nwould therefore be fitting to frame an exception to infringement to account for
\nthe wishes of authors and rights owners (which may not always align with each
\nother), and to ensure that the law would not inadvertently come to either
\nfacilitate unjust enrichment or block technological progress.<\/span><\/p>\n5.1\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0
\n<\/span><\/span>An Exception to Infringement<\/span><\/i><\/h2>\n
\nprimary challenge to contend with is the almost ubiquitous nature of copyright:
\ncopyright has come to protect almost all content, including the most banal,
\nprovided it meets decidedly low thresholds of copyrightability. Although
\ndoctrine somewhat raises the bar through the sc\u00e8nes \u00e0 faire principle and the
\nlike, per the 1957 Copyright Act, sound recordings and cinematograph films must
\nmerely be non-infringing to benefit from copyright protection,[38]<\/b>\u00a0while
\nliterary, dramatic, musical, and artistic works must be original to be
\nprotected by copyright,[39]<\/b>\u00a0with
\njudicial interpretation resulting in \u2018originality\u2019 rarely requiring much beyond
\nhaving a work originate from an author. This has resulted in very little
\ncontemporary content not being protected by copyright.<\/span><\/p>\n
\npossibility would be to have the law treat works differently depending on
\nwhether or not they were commercially available at the instance of authors, and
\non how the relevant AI systems were structured. Although far from a foolproof
\nmode of differentiation, it would seem reasonable to ordinarily assume that
\nsignificant effort had been made to create content legally made available for
\nsale or hire, and it would seem to be fair to exclude such content, including
\ntrade fiction and commercial animation, from the scope of a broad exception to
\ncopyright infringement which allowed copyrighted content to be used as training
\ndata by AI systems.<\/span><\/p>\n
\nit would also seem justifiable to include works which were not commercially
\navailable within the scope of a possible exception to copyright infringement
\nallowing protected content to be used as training data by AI, provided that the
\nworks were not tagged with an explicit indication that they were not intended
\nfor such use, not simply on account of the practical challenges involved in
\ntracing the competent persons to grant licences in respect of all content which
\nwas not commercially available, negotiating licence terms, and paying fees but
\nalso because of the inescapable possibility that many authors may have little
\ndesire to be traced.<\/span><\/p>\n
\nlaw would need to walk a fine line between protecting authors\u2019 revenues and
\ntheir privacy; it is entirely possible that some authors (such as those who
\nhave been subject to gender-based violence in conservative societies and who
\nspeak about their experiences to raise awareness) may desire anonymity for
\ntheir advocacy. Forcibly shredding the privacy of authors at the altar of
\ncopyright maximalism is unlikely to benefit either authors or society at large
\nwhich may ultimately be deprived of their voices.<\/span><\/p>\n5.2\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0
\n<\/span><\/span>Consensual Licensing<\/span><\/i><\/h2>\n
\nit comes to content which is commercially available, too, the law would need to
\ntread carefully noting that within the category lies, on one hand, content
\npublished subject to contractual negotiation which has been made immediately
\navailable for monetisation whether or not its authors have received pecuniary
\nremuneration such as text published by traditional academic and trade
\npublishers, and, on the other hand, content published subject to contractual
\nimposition (including through various software EULAs[40]<\/b>\u00a0and website Terms of Service) which could be monetized but may not have been
\npublished with the intention of having it enter commercial channels such as what
\nis likely to be a significant fraction of end-user posts on social media
\nplatforms.<\/span><\/p>\n
\nthe case of content published subject to contractual negotiation, it is probable
\nthat the identity of authors would be known to publishers while, in the case of
\ncontent published subject to contractual imposition, such knowledge cannot be
\ntaken for granted, and any framework to obtain authors\u2019 consent would therefore
\nhave to be developed cognizant of the need to avoid unmasking the identity of
\nauthors who would prefer enhanced privacy, if not total anonymity. Nonetheless,
\nin both cases, it would be fair to require that authors\u2019 consent for the use of
\ncontent they have created as training data AI be obtained.<\/span><\/p>\n
\nthat publishers acting as content aggregators do not unilaterally license
\ncontent en masse<\/i> to AI companies
\nwithout consent from and remuneration payable to the authors who have created
\nit may require an amendment to the 1957 Copyright Act to ensure that signed
\ngrants of rights to publishers, whether historical or not, are not interpreted
\nto allow such conduct.[41]<\/b>\u00a0Further, while tweaking the provisions in the copyright statute pertaining to
\ncontracts would probably be adequate to obtain informed consent in the case of
\ncontent published subject to contractual negotiation, it would probably be
\nnecessary to include an additional provision either to the statute itself or in
\nsubordinate legislation to ensure that users subject to contractual imposition were
\nin fact aware of the possibility that content they created and uploaded could
\nbe used to train AI, and to give them the option to refuse to allow their
\ncontent to be so used.<\/span><\/p>\n5.3\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0
\n<\/span><\/span>Non-Voluntary Licences<\/span><\/i><\/h2>\n
\nis the question of the manner in which AI systems should benefit from
\nderogations from the usual flow of copyright law whether through exceptions to
\ninfringement, extraordinary contractual perquisites or otherwise.<\/span><\/p>\n
\nat the issue through a utilitarian lens, it could be argued that, to benefit
\nfrom derogations without charge, an AI system should need to be open and have
\nsubscription levels where end-users (being members of the public) could
\nmeaningfully use it without payment. Failing that, if AI systems were some
\ncombination of closed, proprietary and chargeable, there could conceivably be
\nestablished a mechanism through which statutory licences could be obtained in
\nrespect of them to legalize their use of certain forms of protected content as
\ntraining data.<\/span><\/p>\n
\na layered approach is not unknown to copyright law: the provisions in the 1957
\nCopyright Act which help to make copyrighted works accessible to people with
\ndisabilities are contained in an exception to copyright infringement which
\nfocuses on personal, private and not-for-profit access to copyrighted works in
\nstreams independent of the usual commercial channels of trade,[42]<\/b>\u00a0and in a
\ncompulsory licence which allows for-profit operations.[43]<\/b>\u00a0In essence, the law facilitates the possibility of the grant of a licence in
\ncases beyond the purview of the statutory exception to infringement.<\/span><\/p>\n
\nlaw generally does not operate independently: although its substantive
\nprovisions are contained almost exclusively in the 1957 Copyright Act, in
\npractice, its procedural provisions, particularly relating to rights ownership
\nand exploitation, come into their own at the confluence of contract and
\ncopyright law. As a result, it is not exceptional for copyright law to call
\nupon contract law, and there appears to be no reason why it could not do so
\nwith reference to AI. <\/span><\/p>\n
\nproliferation of AI, with creators of all stripes using it, has forced
\ncopyright law to grapple with determining not only the extent to which AI
\nprogrammes themselves should be protected by copyright law but also the extent
\nto which works may be brought into being with technological assistance without
\nsacrificing their eligibility to be protected by copyright (and without their
\ncreators sacrificing the right to be acknowledged as authors).<\/span>\u00a0<\/p>\n
\nthe extent of copyright protection needs to be considered both qualitatively
\nand quantitatively with analyses referring to eligibility for protection and,
\nwhere protection is merited, the duration of such protection respectively. With
\nregard to the latter, it would be prudent to rethink the term for which
\ncopyright protects software in general (within which AI is included), and the
\nterm for which copyright protects the output of AI particularly when it comes
\nto the output of generative AI.<\/span><\/p>\n
\nthe case of AI systems, given the rate at which we now know that technology can
\nprogress, it would seem that lowering the term of copyright for software to a
\nnumber of years comparable to that for which patents or semiconductor
\nintegrated circuits are protected would likely achieve the dual aims of
\nprotecting investment in research, and not supressing further research by
\nallowing those who do not own the rights in specific AI systems to further
\ndevelop them after a certain number of years. Dealing with the term for which
\nthe output of AI should enjoy copyright protection, however, requires one to
\nsimultaneously deal with the issue of whether such output is or should be
\ncopyrightable in the first place.<\/span><\/p>\n
\nthat works meet the criteria to be protected by copyright, to answer the
\nquestion of how long copyright protection should last, it would probably be
\npossible to argue that where works are entirely created by GenAI upon being fed
\nprompts by humans, the duration of protection should be much lower than when works
\nare created by human beings with or without the assistance of technology. This
\nwould divide works into two groups: that which had come into being upon the
\nexecution by AI of instructions from human beings and that which had come into
\nbeing upon the execution of activities by humans possibly with the support of
\nAI; the degree of support would not matter if such a bifurcation were adopted.
\nAt a theoretical level, ignoring the degree of technological assistance
\nemployed to create works would mirror copyright law\u2019s refusal to engage in the
\nqualitative analysis of specific works to determine their copyrightability on
\nthe basis of artistic merit and, on a practical basis, concentrating only on whether
\nor not technological assistance had in fact been employed to bring works into
\nbeing would demonstrate cognisance of the difficulties in assigning a
\nquantitative test to establish the degree of technological support relied upon
\nto create a work without, at the very least, also engaging in qualitative
\nanalyses of the work.<\/span><\/p>\n
\ndifficult as it may be to determine for how long a copyrightable work should be
\nprotected, it is probably even more difficult to determine if a work should be
\nprotected by copyright. While there is no doubt that AI programmes generally
\nmerit copyright protection, the copyrightability of the output of AI programmes
\nis fraught with concerns primarily evolving from two issues: the legality of
\nthe manner in which AI treats content, and the legitimacy of the uses to which
\nhumans put AI with \u2018legitimacy\u2019 being understood to include not just legal but
\nalso ethical considerations. Unfortunately, these issues raise questions which
\ndo not have easily discernible answers: firstly, the question of what exactly
\nAI does with content which has been fed to it as training data, and, secondly,
\nthe question of what exactly humans who use AI are doing with both AI and its
\noutput. Threshing out these two issues is an exercise complicated by the facts
\nthat we do not truly have a workable theory of authorship to guide us, that we
\nare far from certain about the extent to which the framing of prompts amounts
\nto an act of authorship for the purposes of copyright law, and that we rarely
\nseem to know what happens to data within an AI system.<\/span><\/p>\n
\noperations of AI systems tend to be opaque, especially in the case of
\nproprietary AI, and, so, it is often impossible to determine if, in its
\noperation, AI winds up committing copyright infringement or other legal wrongs.
\nIt would, therefore, probably be prudent to frame supplementary guidelines
\ngoverning how legal AI systems may operate, particularly predictive AI systems
\nwhich are used in situations where they could have (sometimes devastating)
\nreal-world consequences for human beings beyond simply creating content that
\nwould be considered artistic; if one takes a step back and looks at the big
\npicture, it is difficult to argue that copyright infringement is the primary wrong
\nthat legal policy should address in this context. For example, AI systems which
\nare used to attempt to accomplish such tasks as determine eligibility for
\nparole, assess the risk of recidivism, or analyse the manner in which an
\nepidemic may spread should, ideally, be rigorously tested so that the risks of
\nemploying AI are well established, and humans are in a position to make
\ninformed decisions factoring in the likelihood of AI output being incorrect.
\nThis may require not only framing policies on how the data inputted into AI
\nsystems may be treated and processed (notably in terms of confidentiality and
\ncopyright) but also framing policies to ensure that AI systems themselves stand
\nup to scrutiny such that claims of their efficacy are not divorced from
\nreality; this may necessitate having the operations of AI systems not operate
\ncompletely with an impenetrable \u2018black box\u2019.<\/span><\/p>\n
\nforcing (often corporate) owners of AI systems to divulge enough about their
\nproprietary software to ensure that those systems actually work as claimed and
\ndo so in a legal manner may require \u2018diluting\u2019 their intellectual property
\nrights, it should be possible to encompass within the copyright framework
\n(which protects software in India) mechanisms beyond those which are extant to
\nprotect information which has been publicly divulged about AI systems (possibly
\nakin to those used by the legal regimes which govern patents and semiconductor
\nintegrated circuits). Ultimately, the law will have to balance proprietary
\nrights and public interest to ensure that one is not privileged over the other.<\/span><\/p>\n
\nfar as the activities of human users are concerned: although it is unlikely
\nthat end-users could be held liable for the operations of AI they did not
\ncontrol, the purpose and manner in which they used AI would likely be factors
\nin determining whether they could be considered the legal authors of works they
\nhad brought into being or, perhaps, had helped bring into being in conjunction
\nwith AI. Complementary to the issue of whether or the output of AI could be
\nconsidered to have been authored by its end-user creator are the questions of,
\nfirstly, whether or not the output could be considered to have violated the
\nrights of people (other than the end-user who had caused the work to come into
\nbeing) such as the right to privacy and the right of publicity, and, secondly,
\nwhether or not the output could be considered to have infringed rights
\nsubsisting in preexisting works.<\/span><\/p>\n
\nsome cases, the issues involved would be relatively easy to clarify. For
\nexample, although copyright law has little enthusiasm to entangle itself with
\nmorality, it is extremely unlikely that a non-consensual, explicit audio-visual
\nwork generated by or with the assistance of computational creativity featuring
\nan identifiable woman would merit copyright protection (not least because it
\nwould have almost certainly been created using preexisting imagery of the
\nwoman); not only would a user who had created such a work not be recognised as
\nan author by copyright law for the purpose of protecting creative content but
\nsuch a user would also likely find themselves being accused of a range of other
\nlegal wrongs including that of defamation which, in India, is a legal wrong
\nunder both tort law and criminal law.<\/span><\/p>\n
\nthe first issue relating to rights violations could be serviceably addressed by
\ndrawing on not just various extant statutes which deal with the issues at play (such
\nas the right to privacy, and the right of publicity) but also by drawing on
\ntort law and, where criminal offences have not been alleged to have been
\ncommitted, contract law to examine if there were any way the lens of \u2018consent\u2019
\ncould justify what would otherwise be unjustifiable rights violations. The
\nsecond issue, that of infringement, would be reasonably easy to address with
\nreference to sound recordings and cinematograph films simply because, under the
\n1957 Copyright Act, infringing works cannot have copyright subsist in them. In
\nthe case of other works, however, the issue could become murky because copyright
\nlaw merely requires non-composite to be original; coupled with the doctrine of
\nindependent creation which allows copyright to subsist in two similar or even
\nidentical works if they have been independently created, it is not impossible
\nthat copyright could subsist in what would, but for the doctrine of independent
\ncreation, have been infringing. Here, too, copyright law is clearly not
\ndesigned to cope with AI.<\/span><\/p>\n
\none hand, while the end-user of a work may imagine that it has been
\nindependently created, the AI itself may have been fed protected content as
\ntraining data. How the law treats the output of AI may have to factor in
\nwhether or not the training data had been acquired legally, and, if so, the
\nterms under which it was so acquired. And, on the other hand, one of the
\nconcerns which has repeatedly presented itself in relation to the output
\ngenerated by AI is that of the legitimacy and copyrightability of works created
\nin the style of contemporary authors. It may be possible to argue that where
\nthose authors have, in any case, commercialised \u2018their\u2019 style such they employ
\nstudios filled with artists to replicate it in a variety of ways, they have
\nalso lost the right to protect their style being imitated by end-user creators
\nusing AI. Such an argument could, by analogy, conceivably apply to the works of
\nany author has made their style formulaic and commercialised although there is currently
\nno body of judicial opinion which could be used to conclusively either
\nestablish or demolish such an argument.<\/span><\/p>\n
\nissue of the who was recognised as the author of works created with the
\nassistance of AI would likely depend not just on the extent to which AI had
\nbeen relied upon but also on the terms of use appurtenant to the AI. Even if
\nthe output were copyrightable, it would not necessarily be possible for a
\ncorporate owner of AI to be considered to be the author (except in cases where
\nthe output was in the form of a sound recording or a film) although it is
\npossible that the company could, through its terms of service, require end-user
\ncreators to assign all rights in the output to the company or to allow the
\ncompany to exploit the output in various stated ways. To avoid situations in
\nwhich companies took undue advantage of end-user creators and, possibly, unjustly
\nenriched themselves at their expense, it would likely be necessary to lay down
\nminimum standards for such contracts in subordinate legislation if not in
\nstatute itself.<\/span><\/p>\n
\nthe use agentic AI alone would likely not result in the creation of a
\ncopyrightable work. However, the use of agentic AI or AI which accomplished
\nspecific and limited tasks in a broader creative process, which allowed the
\nargument to be made that AI had merely supported the creative endeavours of its
\nhuman users, could well result in the creation of works which were
\ncopyrightable and whose creators could be considered to be legal authors. The
\nuse of generative AI would, however, be far less likely to result in either the
\ncreation of copyrightable works or in having human users who prompted the
\ncreation of works be recognised as legal authors.<\/span><\/p>\n
\nline between the two, however, is extremely thin and blurred. For example, AI
\nwhich harmonised a melody could, depending on one\u2019s point of view be considered
\nto be either agentic or generative AI.[44]<\/b>\u00a0Further,
\nthe mere<\/i> harmonisation of a melody, it could be argued, does not merit
\ncopyright protection by itself given that it must, in order to sound sonorous
\nto the human ear, abide by the rules of music theory. This leads to the
\ndevelopment of a situation where it is not obvious who the author of the
\ncontent should be considered to be or, for that matter, whether the content
\nbrought into being is copyrightable at all.<\/span><\/p>\n
\nforms of computational creativity also raise similar concerns. Consider AI used
\nto develop recipes either autonomously or semi-autonomously upon having been
\nfed a range of recipes developed by humans along with rules drawn from hedonic
\npsychophysics, a field which explores what appeals to human beings: although
\nthe AI could be used to autonomously create recipes, it is almost certain to
\nfunction much better in collaboration with human beings who know how to cook
\nthan in isolation.<\/span><\/p>\n