Art Disruption: Art and Technology in the Twenty-first Century

This article is an extract from The Art Law Review, Edition 2. Click here for the full guide

I Introduction

Nowadays, the concept of art is evolving at an extraordinary and exponential speed, and the impact of technology on art is completely changing the concepts of creation, distribution and ownership of artworks.

Artists were among the first to experiment with the 'disruptive technologies', sometimes even delegating most of the creative process to artificial intelligence (AI) or creating artistic autonomous entities that can replicate themselves based on blockchain and smart contracts; at the same time, new technology is also impacting how art is distributed and traded, from the tokenisation of artworks to the creation of art-based cryptocurrencies.

This chapter presents a comprehensive analysis of the impact of new technologies on art and covers significant topics, including the creation of artworks and new art market services.

II Creation

With reference to the relationship between art and technology, it should be first pointed out that the Greek word téchne indicated both the concept of art and that of technology. In our time, on the other hand, we distinguish 'artistic creativity', which primarily has a symbolic and communication value, from 'technical creativity', which primarily has the purpose of solving functional problems. However, in both cases there is a creative act at the origin. Today, this distinction is disappearing with the overwhelming entry/utilisation/exploitation of technology in the process of artistic creation, as well as in the distribution of works, thus returning to a more unitary meaning of téchne.

i Blockchain and smart contracts

To explain how blockchain technology can contribute to the creation of art, we discuss an artwork conceived by Primavera De Filippi,² a software scientist turned artist (and not the only one, as discussed below).

Her artistic project Plantoid was conceived with the goal of showing the potential of technology in creating blockchain-based life forms, meaning independent algorithmic entities capable of sustaining and reproducing themselves autonomously, without human intervention.

How does Plantoid work?³ It is made up of metal sculptures as well as software embedded directly in a blockchain (generating a distributed autonomous organisation). More specifically, it is formed by a series of plant-like metal sculptures, characterised by individual distinctive DNA structures (e.g., rules about their shapes and relationships with donors), associated with a unique digital wallet accepting cryptocurrency from the public. Each plantoid thanks every donation received by moving itself and by playing sounds and lights, in a kind of machine dance. Upon receipt of a predetermined amount of money in a wallet related to a plantoid, a smart contract registered in the blockchain system automatically launches a competition to design a new plantoid, different but based on the same basic DNA. Among all of the projects received from (still human) designers, the donors (bees) select the winning project by sending micro sums to the preferred project. The winning designer then receives the sum in cryptocurrency through a smart contract and he or she can create a new plantoid. And so on. Thus, Plantoid is an autonomous system that could invade the planet with a potentially infinite number of plantoids.

Moreover, the blockchain has recently attracted the attention of the famous Ai Weiwei, who partnered with Irish–US artist Kevin Abosch to create blockchain-based work called Priceless.⁴ Priceless is made up of two standard ERC-20 tokens on the Ethereum blockchain with the name PRCLS and works as follows:

One of the tokens is not available at all, the other is divisible up to 18 decimal places and is to be given away one quintillionth at a time, for free. That is enough for every person on earth. A small amount of the distributable token was instead put into inaccessible digital wallets; the codes for the 12 wallets were printed on paper and sold to art buyers. The codes represent a personal moment in time shared by Ai and Abosch, for example, walking down a street in Berlin, where Ai lives. The pieces of paper are practically worthless, and the codes are a proxy for a fleeting human experience and refer to a valueless token – priceless.⁵

ii AI and algorithmic creation

Other examples of creation through or with the support of AI are two different projects: the works of the French art collective Obvious and the collaborations between Ahmed Elgammal and AICAN.⁶

The first example is the artwork Portrait of Edmond de Belamy, created by a French collective of artists called Obvious, starting with 15,000 portraits dating back to the period between the fourteenth and twentieth centuries and using a generative adversarial network (GAN).⁷ A GAN is a pair of neural networks: the first neural network, which is called generator, randomly mixes all the images that have been put into its memory and creates random combinations; the second, which is called discriminator, selects random results produced by the generator: the ones that better respond to the recurring pattern that had been identified in the original images; the discriminator tries to reproduce human judgement, discarding all the paintings that are not so plausible as potential works created by a human artist. The Portrait of Edmond de Belamy, one of an edition of 10, was auctioned at Christie's in New York on 25 October 2018, and sold for US$432,500 (starting from an estimate of US$7,000 to US$10,000); it was the first AI artwork to be sold through a major auction house.

The use of GANs generates considerable problems in terms of authorship; for example, who is the author of the Portrait of Edmond de Belamy? The basic algorithm that makes the GAN work was developed by Ian Goodfellow (whose name is evoked in the Belamy surname given to the imaginary character portrayed in the painting),⁸ while a second programmer – Robbie Barrat – set the parameters used by the algorithms and loaded the 15,000 ancient portraits into the GAN's memory, making it available online to be freely used by anyone. From there, Obvious slightly modified the software and generated a series of portraits that seem to belong to the same family (named 'the Belamy family'). Finally, Obvious printed the selected images on canvas, gave them a title, and decided to sign them with a portion of the algorithm created by Ian Goodfellow, symbolising the signature of the AI at the base of the entire process. In this case, who is the author of the painting? The basic algorithm (as the signature pretends)? The individual who set the parameters and fed the algorithm with images? Obvious, who chose the images to print, signed them, and gave them a title? Or a combination of all of them?⁹

Some critics argued that this creation by Obvious was nothing more than a reinterpretation of the concept of the objet trouvé or of Marcel Duchamp's readymades or, brilliantly, that the Portrait of Edmond de Belamy could be called 'a machine learning equivalent of a urinal on a plinth'.¹⁰ In other words, the work created by a GAN would be nothing more than a mere material object, which is transformed into a work of art by giving it a title, signing it and putting it into the art circuit.¹¹ After all, in conceptual art, the work of art is not the artefact itself, but it is the artistic action with which the author makes art out of an object, just as Marcel Duchamp took a porcelain urinal, named it Fountain, signed it using the pseudonym 'R Mutt', and then exhibited it in a museum.

Moreover, in early 2019, the HG Contemporary Gallery in New York presented a series of paintings in an exhibition named 'Faceless Portraits Transcending Time',¹² in which the exhibited works were presented as a collaboration between an AI (AICAN) and its creator, Dr Ahmed Elgammal, another professor of computer science turned artist.¹³

The works were based on a set of 3,000 Renaissance portraits. Differently to Obvious, Dr Elgammal used a creative adversarial network (CAN), and not a GAN (in the exhibition labels, the paintings were defined as: 'creative adversarial network print'): Dr Elgammal explained that a CAN is composed of a generator (the same as a GAN) and a second neural network (which we may call a twister) that does not limit itself to judging whether the results conform to the pattern detected in the initial data (such as the discriminator of GANs), but prizes the addition of new elements (deviations) within a given style.¹⁴ In this way, the CAN reproduces the natural evolution of art, which normally does not proceed by radical changes, but through small alterations of a pre-existing style.¹⁵

After the above pioneering works of Obvious and Dr Ahmed Elgammal, an entire artist movement has started, called 'Generative Art'. The opening page of their flagship website states:

Generative Art is a process of algorithmically generating new ideas, forms, shapes, colors or patterns. First, you create rules that provide boundaries for the creation process. Then a computer (or less commonly a human) follows those rules to produce new works. In contrast to traditional artists who may spend days or even months exploring one idea, generative code artists use computers to generate thousands of ideas in milliseconds. Generative artists leverage modern processing power to invent new aesthetics – instructing programs to run within a set of artistic constraints, and guiding the process to a desired result.¹⁶

Thus, neural networks, GANs and CANs have rapidly become a new addition in the twenty-first century artist's toolkit, but people still firmly overlook their output.

iii Art and law in the digital age

Is there a legal notion of what is an artwork? Whereas in many jurisdictions artworks are defined by the medium they are incorporated into (paintings, etchings, sculptures), in others – including Italy – this concept is left open and generally defined by reference to what is recognised as such in the relevant fields, even if this approach is not always seamless.

The huge growth of the forms of expression, the extreme variability of approach to art themes and, lately, the inclusion of disruptive technologies in the creative process, still create many interpretation issues.

Almost a century after Brancusi v. United States¹⁷ – where the Custom Court found that 'while not resembling a bird', the Brancusi sculpture Bird in space was 'beautiful and symmetrical in outline' and, taking into account the emergence of abstractionism, it could be qualified as an artwork and exempted from custom duties¹⁸ – a new tax case concerning art created with disruptive technologies is taking a much less progressive approach. An artist requested an official ruling by the Italian tax authorities, to apply the reduced VAT rate reserved to artworks directly sold by their creator (10 per cent), rather than the ordinary VAT rate (22 per cent). He reported that he digitally created some sculptures on his computer, printed them with a 3D printer and then finalised them by hand. The Italian law, for the application of such reduced tax rate, defines artworks, as far as sculptures are concerned, as 'original works of statuary art or sculptural art, of any material, provided that they are made entirely by the artist; castings of sculptures with a limited edition of eight copies, controlled by the artist or by those entitled . . .'.¹⁹ The tax authorities replied that, for tax purposes, these sculptures cannot be considered artworks as they were neither 'made entirely by the artist' (because the personal intervention of the artist was residual and limited to the final painting of the sculpture) nor were they part of an edition of less than eight cast sculptures.²⁰ This decision of the Italian tax authority could be criticised on several fronts, but mainly on the interpretation of the concept of 'made entirely by the artist': this concept, in fact, must be updated considering the technological tools (such as CAD design or 3D printing) now commonly available to artists. Therefore, while in the past artists used tools such as brushes or chisels, now they use software, algorithms and 3D printers, but this should not make any difference provided that the artists are in control of those tools (as in the case mentioned). Also, it cannot be ignored that the art market is validating these new forms of artistic expression, as, for example, in the aforementioned sale of the Portrait of Edmond de Belamy at Christie's or with the organisation of the Contemporary and Digital Art Fair that took place in New York, and was focused on 'showcasing the diversity of digital artistic mediums, including immersive installation, video art, virtual reality, creative experiments Blockchain-based and more'.²¹

III Non-fungible tokens

'Non-fungible tokens or NFTs are cryptographic assets on blockchain with unique identification codes and metadata that distinguish them from each other. Unlike cryptocurrencies, they cannot be traded or exchanged at equivalency. This differs from fungible tokens like cryptocurrencies, which are identical to each other and, therefore, can be used as a medium for commercial transactions.'²² Therefore, NFTs are digital tokens that have a very specific characterisation and are not replaceable.

NFTs first emerged in the art world in 2017, with a project called CryptoPunks by the US studio LarvaLabs, which freely offered 10,000 unique digital avatars designed by an algorithm. Some had rarer features than others, but all avatars were impossible to copy: NFTs enable digital scarcity, thanks to blockchain technology. This means that digital art for the first time could be subject to the dynamics that naturally drive the art market: context and scarcity. Although they were initially distributed for free, in May 2021 Christie's auctioned a set of nine CryptoPunk NFTs for almost US$17 million (almost double their expected value).²³ Of course, we have to mention the sale of the Everydays: The First 5000 Days NFT by Beeple for US$69 million in March 2021.

In addition, the following announcement on 28 September 2021 is also significant: 'Christie's announces that it has shattered the $100 million ceiling in NFT sales worldwide. Pushing Christie's over this historic mark was the record-setting success of Christie's first ever NFT auction in Asia, which closed online in Hong Kong on 28 September and achieved a total of HK$122 Million (USD $15.6M).'²⁴

From these impressive figures, it is possible to understand the growing importance of the NFT phenomenon in the art market; but what makes them so attractive?

To answer that question, it is necessary to understand how NFTs work and, specifically, how they apply to art. By purchasing an NFT, you are purchasing a token and the artwork attached to it, which is not normally stored on the token (because of the prohibitive cost of minting large amounts of data). The NFTs always contain ownership data as well as metadata related to relevant digital artwork, plus a pointer to where the artwork is stored (i.e., a URL or an address of the interplanetary filing system): this is called off-chain storage. Each transaction concerning the NFT (starting from its creation, called minting, to each subsequent transfer) is recorded on the blockchain: in this manner, no one can modify the record of ownership or copy and paste a new identical NFT (whereas, the content stored off-chain can be changed or removed). This functioning of NFTs, when applied to art, ensures digital scarcity by allowing the creation of unique pieces or limited and numbered editions. The majority of art related to NFTs reside on Ethereum, a totally decentralised, open-source blockchain with smart contract functionality.²⁵ That is why an NFT can be considered the digital version of a certificate of authenticity, embodied in the blockchain.

As is the case with any new disruptive technology, there are pros and cons to NFTs.

As to the pros, if you own an NFT:

1. you can easily prove you own it;
2. no one can manipulate it in any way;
3. you can easily sell it; or
4. you can hold it forever, resting comfortably knowing your asset is secure in your wallet.

From a different perspective, if you are an artist and you create an NFT:

1. you can easily prove you are the creator;
2. you determine the scarcity;
3. you may earn royalties every time it is resold (if this is provided for in a smart contract embedded in the NFT itself); and
4. you can sell it on any NFT market or peer to peer.

However, the risks associated with the 'minting' of NFTs cannot be ignored. One of the problems is that anyone can mint an NFT, creating a file and turning it into an NFT, even if they have no right to the content of the NFT. Consequently, someone may mint an artwork by another artist causing an intellectual property infringement, which is very difficult to fight because of the anonymity permitted by NFTs.²⁶ Additionally, much criticism is emerging about the amount of energy required to mint an NFT or transfer it when using the 'proof of work' protocol (as for NFTs minted on Ethereum); alternative platforms using the much more energy efficient 'proof of stake' protocol (such as Tezos) are now emerging.

IV Post-covid art market

While the digital transition of the art market has been occurring for a few years, the lockdowns imposed by the covid-19 pandemic have dramatically accelerated it in 2020. A recent Art Basel survey concerning the first six months of 2020 reveals that the share of online sales of art galleries 'rose from 10% of total sales in 2019 to 37% in the first half of 2020'.²⁷

Therefore, the presence of digital tools in the art market is no longer marginal, but absolutely central. The new technologies are in fact impacting fundamental aspects such as the assessment and guarantee of authenticity and provenance, the legal title of the seller, the opportunities of fractional ownership, and the way proceeds from art sales are shared among the various players involved.

iEvidencing authenticity and provenance

The basic issue involved in art sales is the authenticity of the work of art and this aspect has recently become quite problematic when many authentication committees (Warhol, Basquiat, Haring, to name a few), facing a growing number of disputes with disgruntled collectors, have ceased to issue certificates attesting the authenticity of the same, thus creating a confusing grey area in which collectors do not know who to turn to.²⁸

One way to solve this problem is to create digital incorruptible evidence of the chain of title of an artwork, best of all from its creation.

One example of this is the activity of a US company named Verisart, which applies blockchain technology to combine transparency, anonymity and security to protect records of creation and ownership of artworks and collectibles. On the home page of the Verisart website, it is stated that:

Verisart is building evidentiary infrastructure for artworks and collectibles that is verifiable by anyone. . . . Records are encrypted and timestamped by the world's most-trusted decentralized ledger. Certificates are easy to manage and can be shared or transferred at any time.²⁹

The identification of the work is actually ensured by Verisart by uploading a high-definition photo of the artwork, which makes it easy to spot any potential forgery through image-recognition technology.

Needless to say, the effectiveness of the system depends on the accuracy of the data initially entered. To demonstrate and emphasise this, Terence Eden, a technology enthusiast and influencer, uploaded on Verisart, as his own work, Da Vinci's Mona Lisa, thus generating a blockchain provenance record that claims he is the true 'author and owner' of Mona Lisa; given the (almost) impossibility of erasing the 'blocks' of a blockchain, this information will remain recorded in the system forever, although obviously no one will realistically believe it to be true.

ii Artwork recognition and stroke recognition

Technology provides the art market with a new powerful tool to verify the authenticity of artworks: AI, applied either to the artwork as a whole or to characterising elements thereof, such as the strokes in a drawing.

An interesting project concerns the verification of the authenticity of drawings. Rutgers University, with a team led by Dr Ahmed Elgammal³⁰ and the Atelier for Restoration & Research of Paintings,³¹ have commenced a research project on several hundred drawings by a few famous artists, trying to show how each line drawn on a sheet of paper is practically comparable to a fingerprint of the artist. In particular, the algorithm they created broke down almost 300 line drawings by Picasso, Matisse, Modigliani, and other famous artists, into 80,000 individual strokes. Then, a recurrent neural network learned which features in the strokes were important to identify the artist and, at the same time, the researchers also trained a machine-learning algorithm to look for specific features, such as the shape of the line in a stroke. With both algorithms working in tandem, the researchers were able to correctly identify artists around 80 per cent of the time. The same technology is able to distinguish an original drawing from a certain artist from a forgery of the same.

v Art market services

The new technologies have also permitted the emergence of a number of new services directed towards the art market, a few examples of which are discussed below.

iArt appraisal

The soft-spoken world of art appraisers is now facing competition from an AI-based appraisals system. An Italian start-up, Kellify.com, has created an AI-powered methodology that, by processing art market big data, is able to provide automated appraisals of artworks, as well as forecasts of their future market trends. Being able to provide real-time detailed objective insights and forecasts, remotely, and 24/7, Kellify immediately attracted the attention of art insurers, art lending institutions and art funds.

ii Instant access to artworks data

The Magnus app is an innovation that could be considered the Shazam of the art world.³² This app makes it possible, on the basis of a photographic reproduction of an artwork, to search within the databases of all the most important auctions and all the other available data on the art market and to instantly find data on the work and on its previous commercialisation (author, year of creation, sales history, or if and when it remained unsold). Therefore, Magnus allows anyone to have instant access to such information and, as a consequence, leads to greater transparency in the art market. The app creates a real disintermediation of the relationship between collector and works of art: the collector no longer needs an expert to access this data.

iii Marketing intelligence

Thread Genius, established in 2015 and then acquired by Sotheby's in 2018, through a set of AI algorithms, carries out market intelligence activities, endeavouring to identify which works might be of interest to a collector on the basis of his or her previous acquisitions. Similar AI-based services are now offered by Artrendex through its ArtPI platform,³³ which enables museums, galleries and fairs to better engage with their visitors and clients by offering them images of artworks that have characteristics that correspond to other artworks they appeared to admire.

iv Shipping services

ARTA is a digital platform delivering global logistics and services for art shipments.³⁴ It offers automated quotes for packing, shipping and installing artworks in a few minutes, which is a significant improvement on the hours or days normally needed to get such an estimate from a traditional art shipper.

v Art curation

After the Swiss collective fabric|ch created an artwork for the 2019 exhibition 'Entangled Realities: Living With Artificial Intelligence' at Basel's House of Electronic Arts, which was entitled Atomized (curatorial) Functioning and that 'leveraged algorithms to produce a steady stream of layout variations for the very show in which it appeared',³⁵ we now have the first biennale entirely curated by a robot: the chief curator of the 2022 Bucharest Biennial will be Jarvis, an AI program in development from the Vienna-based studio Spinnwerk that will 'use deep learning in order to learn by itself from databases from universities, galleries, or art centers' and select works that fit the chosen theme. As to the impact of digital technologies on art, it is also interesting to underline that this entire biennale will only be available in virtual reality, through a virtual reality headset.³⁶

vi Sales

New technologies are also creating new opportunities for how artworks can be traded.

Experiments in fractional ownership offered through online platforms may involve either traditional co-ownership contracts (see Art Share)³⁷ or create digital ownership certificates through the blockchain, creating digital tokens, which is known as 'tokenisation' of artworks.

This business model based on blockchain technology was first created in 2017 by Maecenas.co, a blockchain-based platform that allows anyone to buy, sell and trade fractional ownership in masterpieces on a liquid exchange.³⁸ Fractional ownership then permits even small collectors to invest their money in (fractions of) important works of well-known artists, rather than in works of lesser known artists, thus increasing their chance of having access to the most dynamic and liquid portion of the market. Its mechanism also permits collectors who owns important works to sell only a portion of their value (less than 40 per cent), cashing the revenues of the fraction sale and maintaining possession of the work, thus making the artwork 'fractionally liquid'.

On 6 September 2018, Maecenas sold, through tokenisation, more than 30 per cent of Andy Warhol's painting 4 Small Electric Chairs (1980), for US$1.7 million: the tokens acquired can be traded through the Maecenas exchange.³⁹

This new experience and the rapidity with which it spread among investors has pioneered the massive implementation of this technology in the art market, demonstrating the merits of developing and improving the market through technology.

Now other players are entering the same art tokenisation market: among the most original is Snark.art, which sold 2,304 tokens (called atoms) of a video work called 89 Moments Atomized by Eve Sussman, the sale involving not only a digital token but also ownership of a fraction of the screenshot of the same video: namely, each atom is 400 pixels of the entire visual frame with an approximately 10-minute duration (that is the duration of the full video), and the mechanics of blockchain ensure that each atom is uniquely assigned to the owner, so that provenance and singularity cannot be pirated or faked.⁴⁰

Tokenisation is proving particularly attractive to a young and digital generation of buyers and is now being applied to all sorts of collectibles (thus, not only fine art, but also sport memorabilia, racing cars, watches, comic books), there being at least 10 other companies offering tokenisation opportunities.⁴¹

A totally different experience is that of the online platform dada.nyc, which only trades in digital art. This website does not offer fractional ownership, but each sale of a digital work (normally offered in numbered editions) is actually recorded on a blockchain and this guarantees the 'digital scarcity' as only one work with a certain edition number will exist and be recorded on the Ethereum blockchain. Moreover, the platform has implemented a unique way of sharing the proceeds of each subsequent sale of a digital artwork traded through it, with only 60 per cent of the proceeds going to the seller, 30 per cent going to the artist and 10 per cent to the platform, with all payments being made automatically through smart contracts.⁴²

VII Conclusion

Disruptive technologies (such as blockchain, AI, neural networks, smart contracts, virtual reality and augmented reality) are creating entirely new works of art and business models, which are revolutionising the quite steady art market.

As most of these innovations are sailing in unknown waters, entirely new legal issues are emerging as to authorship, authenticity, representation and warranties, financial regulation of art trade, cross-border sales, transactions' traceability and privacy, and use of smart contracts.

Therefore, the art law lawyer of the future must also, inevitably, be a technology lawyer.

---

Footnotes

¹ Massimo Sterpi is a partner at Gianni & Origoni.

² Primavera De Filippi is a permanent researcher at the National Center of Scientific Research in Paris, a faculty associate at the Berkman Klein Center for Internet & Society at Harvard University and a visiting fellow at the Robert Schuman Centre for Advanced Studies at the European University Institute.

³ See www.okhaos.com/plantoids/.

⁴ See www.vice.com/en/article/qvmm9m/ai-weiwei-kevin-abosch-blockchain-art-priceless.

⁵ See www.thepeakmagazine.com.sg/lifestyle/cryptocurrency-prices-artwork-maecenas-dadiani-syndicate/.

⁶ Meaning 'artificial intelligence creative adversarial network'.

⁷ 'Obvious is a collective of researchers, artists, and friends, working with the latest models of deep learning to explore the creative potential of artificial intelligence. They are behind the sale of the first AI artwork to go through a major auction house. They use their work to share their vision of artificial intelligence and its implementation in our society' (www.obvious-art.com/).

⁸ Goodfellow almost translates as bel amie in French.

⁹ For an analysis on the IP implications of Portrait of Edmond Belamy, see www.ipkitten.blogspot.com and www.moc.media/en/2527. The following is also reported by one of the co-founders of the Obvious collective, Hugo Caselles-Dupré: 'If the artist is the one that creates the image, then that would be the machine. If the artist is the one that holds the vision and wants to share the message, then that would be us.'

¹⁰ See www.theatlantic.com/technology/ archive/2019/03/ai-created-art-invades-chelsea-gallery-scene/584134/.

¹¹ As Mario Klingemann, a German artist who has won awards for his own work with GANs, commented, 'I wonder why [Obvious] missed the opportunity to declare their work as an AI-readymade and bring us the first digital Duchamp.'

¹² See www.hgcontemporary.com/exhibitions/faceless-portraits-transcending-time.

¹³ Dr Ahmed Elgammal is a professor at the Department of Computer Science, Rutgers University. He is the founder and director of the Art and Artificial Intelligence Laboratory at Rutgers. He is also an Executive Council Faculty at Rutgers University Center for Cognitive Science. Dr Elgammal is the founder and CEO of Artrendex, a start-up that builds innovative AI technology for the art market.

¹⁴ The CAN is described in a 2017 paper by Ahmed Elgammal: 'The system generates art by looking at art and learning about style; and becomes creative by increasing the arousal potential of the generated art by deviating from the learned styles' (CAN: Creative Adversarial Networks, Generating 'Art' by Learning About Styles and Deviating from Style Norms, Ahmed Elgammal et al., June 2017, www.researchgate.net/publication/317823071_CAN).

¹⁵ See Zack Thoutt, 'What are Creative Adversarial Networks (CANs)?', https://hackernoon.com/what-are-creative-adversarial-networks-cans-bb81d09aa235.

¹⁶ https://aiartists.org/generative-art-design.

¹⁷ Brancusi v. United States, 54 Treas. Dec. 428, 429 (Cust. Ct. 1928).

¹⁸ See www.thelegalpalette.com/home/2018/3/20/brancusis-bird-in-space-is-it-a-bird-or-is-it-art#:~:text=In%20the%201928%20case%20Brancusi,for%20the%20free%20import%20duty.&text=United%20States%20to%20transform%20the,into%20a%20more%20contemporary%20standard.

¹⁹ Decree Law No. 41/1995.

²⁰ See www.agenziaentrate.gov.it/portale/documents/20143/2665720/Risposta+all%27interpello+n.+303+del+2020.pdf/907a73a3-a850-4849-8907-4630065901b5.

²¹ See https://cadaf.art/. 'CADAF is a discovery and interaction platform dedicated to digital and new media art. It regularly hosts international art fairs, special events, art exhibitions and talks. Discover and connect with the world's most important artists, galleries, curators and collectors of the digital art market. CADAF is a product or New Art Group LLC, along with New Art Academy, dedicated to art + tech education.'

²² See: https://www.investopedia.com/non-fungible-tokens-nft-5115211.

²³ See: https://markets.businessinsider.com/currencies/news/cryptopunk-nft-christies-auction-value-crypto-art-larvalabs-2021-5-1030419403.

²⁴ www.christies.com/about-us/press-archive/details?PressReleaseID=10210.

²⁵ See: https://ethereum.org/en/. NFT standards also exist on other blockchains, such as Bitcoin Cash and Flow. This happens because Ethereum is very expensive for storing large files.

²⁶ For example: (1) the artist Weird Undead found her tweet (with an embedded artwork) up for auction on NFT marketplace platform OpenSea; after filing legal notices with OpenSea, citing what she called 'insane and pointless copyright infringement', the token vanished from the site; (2) Corbin Rainbolt – a paleo-artist from Indiana, US, after realising that someone linked an NFT to his tweet featuring his Spinosaurus painting, had to delete all his art tweets from Twitter and repost them with watermarks; and (3) visual artist Rosa Menkman discovered four of her artworks had been put up for auction on OpenSea – she was also able to have her work removed from the marketplace platform.

²⁷ See www.artbasel.com/about/initiatives/the-art-market.

²⁸ Inter alia, see, for example: (1) the Andy Warhol Foundation, which will no longer authenticate the works of the well-known pop art artist; (2) the Authentication Committee of the Estate of Jean-Michel Basquiat, which, after having attributed more than 2,000 authenticated works over the course of 18 years, completely ceased to authenticate the works of the US artist in September 2012; and (3) the Keith Haring Foundation, which has dismissed the authenticity committees and no longer examines works attributed to the artist.

²⁹ See www.verisart.com: 'Verisart is building evidentiary infrastructure for artworks and collectibles that is verifiable by anyone.'

³⁰ See footnote 13.

³¹ See www.arrs.nl/arrs_en.html: the Atelier for Restoration & Research of Paintings, founded in 1991, 'has been consulted for professional restoration, authenticity research, expertise and condition reports by national museums, government agencies, institutions and auction houses. Our international client base includes leading art dealers and private collectors worldwide.'

³² See www.magnus.net/about/: 'Our mission is to make the art market transparent.'

³³ See www.artpi.co/.

³⁴ See www.shiparta.com/.

³⁵ See https://news.artnet.com/market/ai-art-business-intelligence-report-2020-1812288.

³⁶ See https://news.artnet.com/exhibitions/bucharest-biennial-curated-by-artificial-intelligence-1872342.

³⁷ See www.artsharesales.com/.

³⁸ See www.maecenas.co/: Maecenas, the art investment platform.

³⁹ www.finivi.com/tokenized-art-alternative-investment/.

⁴⁰ See https://medium.com/@snarkdotart/snark-art-launches-new-art-and-blockchain-laboratory-with-acclaimed-artist-eve-sussman-press-6e7777b50142.

⁴¹ See www.bloomberg.com/news/articles/2020-02-07/you-can-own-a-fraction-of-a-warhol-but-should-you.

⁴² See www.artmarket.guru/le-journal/blockchain/yehudit-mam/.