Plagiarism is not a new evil and has been facilitated in the past through ghost-writers, copy and paste activities from different sources (books, internet, etc) and other means. However, plagiarism software has been our true saviour and over the last decade academia has benefitted greatly by anti-plagiarism technology. Alas, the news now is that technology has turned the tables on us, and the new villains are Artificial Intelligence (AI) powered bots such as ChatGPT. This tool is powered by the vast knowledge available on the internet and offers conversational responses. Behold, this has triggered an avalanche of cries from worried academics and universities, with many bemoaning the acute possibility of all our students becoming notorious (and successful) cheats. Many reports have been printed about how the output of ChatGPT has already fooled various universities and public examination authorities (in the fields of law, medicine, etc). Such an approach is of course alarmist and hopefully, once the hysteria dies down, we all can take a calm look at the facts as they present themselves. After all, despite all the hype, self-driving cars are still a niche market and the less said about the so-called 3 D Printing revolution, the better.

Under intellectual property laws, only a human being can be regarded as the author or inventor. Thus, for example, according to the European Patent Office in the DABUS patent case, AI generated inventions are patentable only if a natural person is named as the inventor. Similarly, in the “monkey-selfie” copyright dispute, it was asserted in USA that a monkey, being a non-human, cannot own copyright under US laws. So any plagiarism, conducted by a student, with the help of ChatGPT would still be her own work, for which she alone would be responsible. ChatGPT is just a tool, a bit better than those preceding it, and certainly not the last of its kind.

Here is a short assessment of possible solutions in this regard:    

1. ChatGPT styled AI powered bots are BANNED – This is extremely unlikely. Although ChatGPT uses material from online resources and this could potentially violate copyright in those resources, one must remember that the US Supreme Court in the Sony “Betamax” case in the 1980s, held that if the technology in question has significant non-infringing uses, then the rights of others to engage in legitimate and unobjectionable commercial purposes cannot be denied. This legal philosophy has been adopted very widely through out the world when interpreting copyright issues in the digital sphere, except when there is actual infringement of intellectual property rights for malafide purposes (such as the Napster case, Pirate Bay case, etc). 
2. ChatGPT styled AI powered bots use WATERMARKs to identify the end product as being AI generated – this depends on the development and adoption of industry standards, best practices, legislative requirements, etc.
3. Development of ANTI PLAGIARISM software which detects AI generated works. This is already possible as there are effective AI content detectors that are available (Coalition Technologies, Copyleaks, etc). It is just a matter of time before plagiarism software develops to counter ChatGPT.
4. Keeping aside technology, academic tutors can still revert to the OLD SCHOOL assessment methods such as closed book exams in classrooms, hand written exam answers, oral exams, lengthy oral defences (in person) of theses, etc.

As the famous French writer Jean-Baptiste Alphonse Karr wrote in 1849  “plus ça change, plus c'est la même chose” – the more things change, the more they stay the same.

Read more about our Law Master's programme: https://taltech.ee/en/masters-programmes/law