Who wrote this? The crisis of authorship — Vedas, Shakespeare and now, AI

For centuries, people have looked to an author’s name to understand who created a work, who deserves credit and who should be held responsible for its ideas and mistakes. Artificial intelligence is now challenging that familiar understanding of authorship as AI tools become increasingly involved in writing, research, software development, science, journalism and other professional fields. Universities are struggling to determine whether students are using AI to complete work that is supposed to demonstrate their own abilities, while publishers and researchers are facing similar questions about how much machine assistance is acceptable. The debate is often framed around a simple question: Was AI used? But that may not be the most useful way to understand the problem. The more important issue may be whether a human being remains responsible for the ideas, judgment, accuracy and meaning of the final work. The publishing industry has already encountered this challenge, with questions emerging over whether writers can prove that manuscripts were produced primarily through their own work. Such cases demonstrate how traditional ideas of authorship are becoming harder to apply in an era when AI can generate and transform large amounts of text. Yet the idea of authorship has never been as straightforward as it appears. Shakespeare, for example, is traditionally regarded as the sole author of his famous plays, but scholars have long debated the extent of collaboration involved in several works. Collaboration was common in the theater of Shakespeare’s era, meaning that the modern image of the solitary literary genius may not accurately reflect how many important works were created. The Vedas provide a very different example, as traditional Hindu thought regards them as apauruṣeya, or not of human authorship, with ancient rishis understood as receiving and transmitting knowledge rather than claiming ownership of it. These examples suggest that human knowledge has never depended entirely on identifying a single individual responsible for every word. Language, science, culture and literature are all built collectively, with every generation inheriting ideas and knowledge developed by those who came before. Artificial intelligence is making this collective nature of creation even more visible because machines can now participate in activities once associated almost exclusively with human intellectual work. AI can summarize research, analyze documents, generate computer code, suggest arguments, improve writing and create images, music and other forms of content. The challenge lies in determining when AI is enhancing human ability and when it is replacing the ability being evaluated. If a student uses AI to correct grammar, improve clarity or identify weaknesses in an argument, the technology may function like an advanced editor or intellectual assistant. But if the student allows AI to develop the argument, conduct the analysis, write the paper and produce the conclusions, it becomes difficult to claim that the final work represents independent thinking. There is no simple dividing line between these situations because AI assistance exists along a broad continuum. The question is therefore less about whether a machine was involved and more about what role the machine played. This distinction is becoming particularly important in scientific publishing, where many journals require disclosure of generative AI use while refusing to recognize AI systems as authors because machines cannot accept responsibility for the accuracy or consequences of their work. Authorship has traditionally involved more than producing words; it also involves standing behind what those words claim. Another concern is that widespread AI use could weaken individual voice. AI can make writing more polished and organized, but it can also make different people’s writing increasingly similar. If millions of users rely on similar systems to produce essays, speeches, proposals and opinions, communication may become technically sophisticated while losing some of the distinctive qualities that make human expression meaningful. The deeper issue, therefore, may not be whether AI threatens authorship but whether people continue to develop their own judgment and voice while using it. The same principle becomes even more important when AI systems begin influencing decisions rather than simply producing text. If an AI system contributes to a medical decision that harms a patient or an autonomous vehicle causes a fatal accident, the machine cannot accept moral responsibility in the human sense. Responsibility must ultimately remain with the individuals and institutions that design, deploy, supervise or approve these systems. This suggests that the future of authorship may increasingly revolve around responsibility rather than the physical production of words. Instead of asking only who wrote something, society may need to ask who conceived the idea, who exercised judgment, who verified the information, who approved the final result and who is prepared to accept the consequences if it proves wrong. Shakespeare demonstrates that uncertainty about individual contribution does not necessarily destroy the value of a work, while the tradition surrounding the Vedas shows that knowledge can be meaningful even without modern notions of individual ownership. AI is now forcing society to reconsider these ideas in a technological environment where machines can actively participate in creating knowledge. The value of human contribution may increasingly lie not in manually producing every sentence but in deciding what should be produced, determining whether it is accurate and meaningful, and taking responsibility for the final result. In the age of AI, the most important question may therefore no longer be simply “Who wrote this?” but “Who is willing to stand behind it?” The person who deserves the greatest credit may ultimately be the one who exercises the most judgment and remains willing to accept responsibility when the technology gets something wrong.