Neurotecnología e Inteligencia Artificial: Fronteras legales y de gobernanza

129 approximately 550 milliseconds (i.e., 0.5 seconds), and a person’s conscious realiza- tion of an intention. Extrapolating from this, a BCI-gAI combination might anticipa- te a creative idea emerging in the mind and start rendering it into an artistic medium almost instantaneously. This scenario is grounded in existing trends: as resolution and accuracy of brain-scanning improve, and as gAI models for text, image, audio and video generation become better, the gap between thought and creation conti- nues to narrow, while the gap between thought processes and the realization of a thought remains stable. From a legal perspective, such direct brain-to-art (B2A) technology challenges the very definitions of authorship and fixation. Copyright law traditionally requi- res that a work be “fixed in a tangible medium of expression”, 6 by or under the authority of an author. In a B2A procedure, is the thought itself the work, or is it only a work once the output file is generated by the AI? The idea is that each work begins with a conception in the mind, which only gains copyright protection once expressed. Here, the expression manifests itself via a machine interpreting neural signals. One might argue the machine is simply acting as a translation tool rea- ding brainwaves of the human thinker, who could still be considered author, since the originating creative act, having an idea and wanting to express it in a certain idiosyncratic way, is human. This view can be described as Cartesian dualism: the human mind is a “thinking thing,” ( res cogitans ) and is the source of creativity; while the machine is an “extended thing,” ( res extensa ) executing the physical expression of that creative will. In this view, the BCI is “a genuine prosthetic of the brain,” (Vidal, 1973) in particular the artist’s mind. The law could accommodate this by treating the human as the author. Some scholars have likened such gAI tools to the “next step” in a long history of technology extending human creative powers, no different in principle from the introduction of photography or digital editing software, technologies that initially stirred debate but were ultimately absorbed into the legal concept of authorship, where the human operator remained the author. The opposing view, however, is that when gAI becomes deeply involved, if it begins to augment or modify the human’s idea substantially, the human’s claim to authorship weakens. What if the gAI injects its own elements that the human did not specifically imagine? For example, suppose an artist merely thinks of a vague concept, “a surreal landscape with emotional reso- nance,” and the gAI, drawing on its training data, fleshes it out into a complex pain- ting with many details that the person never envisioned. Is that person the author of those details, or are they essentially generated by the AI, and thus, by nobody, if the gAI cannot hold copyright? Some experts forecast scenarios where creative content is a collaboration between human minds and gAI systems to such an extent that pinpointing a single author becomes legally and philosophically fraught. 6 17 U.S. Code § 102(a). The entanglement of BCI, gAI, ART and IP