We address emotionalized artificial intelligence (EAI) in a semiotic interrogation, by reflecting on how emerging types of human-computer interaction are changing human environments. We mean to lay the foundations for a semiotic approach to AI, which so far lacks, despite the semiotic interest for cognition and interaction. Affective states are not just embodied and embedded, but also extended and enacted by way of extra-psycho/corporeal processes and (im)material artifacts. Fine art, film, music, books, drugs, computers and photographs enable emotional-cognitive responses. Proponents of situated affectivity contend that some artifacts can be experienced as an integral part of the self to the point where they shape subjectivity. While generative AI and Large Language Models have catapulted intelligent machines to the forefront of public attention, EAI remains little-known. This technology involves computer vision, deep learning algorithms, big data, natural word processing, voice tone analytics, advanced biometric sensors and actuators. EAI can adapt its behavior and tailor responses to suit the emotional context of human interaction. Still in a nascent stage, affect-recognition algorithms have many uses, from healthcare to entertainment, advertising, and surveillance. We ask what happens when an affective artifact transforms into an intelligent machine capable of inventorying, inferring, simulating and regulating a person’s affective life?How does the discourse on situated affectivity change when artifacts become a Lacanian mirror, and humans delegate emotional life experience to machines? These ponderings call into question not just the role of algorithms in human thought and feeling but of semiotic agency, broadly.