Artificial intelligence: some legal and ethical issues
Hank Ze
The surge of development in artificial intelligence in recent years has brought numerous opportunities for humankind. One particular type of AI – generative AI – has revolutionized the field through its extraordinary ability to generate precise answers written in human-like style from minimal user input. However, such innovations also unequivocally come with dangers, including academic dishonesty and, in more severe cases, potential criminal charges. Currently, due to the embryonic stage of AI, these issues are by no means clear cut. For ease, I will focus solely on UK legislation. In order to analyse whether the contemplation of submitting an AI- written essay is ‘ordinary attempt ed theft’, it must be the case that the successful act of submitting an AI - written essay is ‘ordinary theft’: an attempted crime can only be charged if the successful act is a crime in itself. This consequently introduces the Theft Act 1968. 1 The Act stipulates under Section 1 that ‘ A person is guilty of theft if he dishonestly appropriates property belonging to another with the intention of permanently depriving the other of it ’ . There are five parts to this definition: 1) ‘ Dishonestly ’ – the person had no right in law nor consent to deprive the other of it; 2) ‘ Appropriates ’ – the person had taken on the rights of an owner, innocently or not; 3) ‘ Property ’ – all real and personal property, including intangible property; 4) ‘ Belonging to another ’ – the victim had possession or control over the property; and 5) ‘ with the intention of permanently depriving the other of it ’ – a person who appropriates the property without intending to permanently deprive is nevertheless deemed to have such an intention. Altogether, all five parts must be satisfied in order for a person to be guilty of theft. It seems that the possibility of committing theft when submitting an AI-written essay is very much associated with the nature of content creation by generative AI. Recently, for example, there have been headlines on the notion of AI ‘ stealing from artists ’ due to artwork being generated based on styles of particular artists. 2 Firstly, to say that AI commits theft is incorrect. AI uses complex algorithms such as natural language processing (NLP) and generation (NLG) 3 to collate data from a variety of sources, then paraphrases it to produce unique and original work. In this sense, AI used for writing essays is merely a rewriting tool. 4 Therefore, this action does not satisfy parts 1) and 5) of the definition of theft, because the person using the AI (henceforth, P) has not deprived the original authors of anything; part 2) also fails to stand, because P has not yet presented the AI-written essay as their own work. Therefore, using AI to write an essay is not theft.
By this reasoning, nor does the submission of an AI-written essay constitute theft. It is certainly true that, in this case, part 2) does stand, since P has presented the essay as their own work, and therefore as
1 Theft Act 1968 (c.60). 2 Chayka, ‘ Is A.I. Art Stealing from Artists? ’
3 Spooner, ‘ AI Content Generation: Pros and Cons of AI Content Writing ’. 4 Eaton, ‘ Artificial Intelligence and Academic Integrity, Post-Plagiarism ’ .
207
Made with FlippingBook - PDF hosting