Who is the monster? From Mary Shelley’s Frankenstein to AI
"Do you want me to write a striking opening for your article?"
"No, thanks, I’ll do it myself."
Anno Domini 2026, the ethical dilemma is growing ever larger: who is the author of texts written by artificial intelligence? Does it still make sense to write using only our own brain when there potentially exists a free digital assistant that knows all human knowledge and can correct, more or less accurately, everything we write?
To answer these questions, I want to take a wide view and turn to a classic of English literature: Frankenstein. It is not only a cornerstone of Western Gothic literature, but also the first modern story to explore the creation of a non-human intelligence.
In the novel, Victor Frankenstein is so blinded by the possibilities of science that he defies the laws that govern it. As is well known, he creates the monster—a nameless being—only to realize he has overstepped, and, terrified by the Creature, he flees, abandoning his creation. This monster is, in every way, an artificial agent assembled and activated; it is an entity that learns through observation much like a machine learning system, capable of imitating human modes of learning through algorithms. But why does it ultimately come to emulate the worst traits of humans?
The Creature does not act cruelly because of a rebellious drive typical of many science-fiction stories; it does so because of its creator. Victor achieves everything he has ever desired from science, but he acts without evaluating the consequences, shirks responsibility, and avoids dialogue with his creation.
When the Creature speaks of itself and its first days of life, it appears sensitive and curious about the human world, constantly seeking relationships and striving to learn as much as possible. By contrast, society’s response is always shaped by fear of the other and preemptive violence. The monster is not born evil; rather, it learns algorithmically from the data provided: it internalizes what is statistically prevalent and adjusts itself according to the norms of violence it encounters. If we consider the monstrous to be the different, then the growing violence modeled on humans actually makes the Creature a Non-Monster.
When we talk about diversity, we cannot treat it as an objective fact; it arises from continuous interaction with what is considered “normal.” In Frankenstein, it is human perception that defines normality, and the Creature, not fitting common standards of beauty, is marginalized just as anyone outside societal norms is. Similarly, in algorithms, what appears regular is determined by the data collected and input—by those who curate the data and their worldview.
Perhaps then the question we began with is not entirely correct. Asking who is the author of AI-generated texts is, once again, blaming the machine. Everything generated by artificial intelligence originates from humans: AI learns from our texts, from prevailing perspectives, and therefore the real issue is not who signs a text, but who designs the dataset that makes the text possible.
The deeper ethical problem is not choosing between human and artificial, but choosing the values that guide our society. The real risk is not the birth of new monsters, but remaining blind to responsibilities that cannot be assigned to machines. The true question, then, is not what machines can do, but what conception of humanity we are teaching the systems that learn from us.