AI-generated Shakespeare: A corpus stylistic analysis of ChatGPT-4's generated and adapted scenes

Abstract

Generative AI is reshaping literary production, raising critical questions about textual integrity and authenticity. The rise of generative AI challenges foundational concepts in literary theory, including authorship, stylistic integrity, and the very ontology of the canonical text. This study uses the case of Shakespearean adaptation by GPT-4 to interrogate not merely the capability of AI, but its function as a cultural agent that reframes literary heritage through the logics of accessibility, simplification, and algorithmic bias. A corpus stylistic and comparative textual analysis is conducted on key scenes from Romeo and Juliet, The Merchant of Venice, The Winter's Tale, King Lear, Hamlet, and Othello. The study evaluates AI-generated adaptations of the original texts, focusing on linguistic, stylistic, thematic, and contextual fidelity. Findings indicate that while GPT-4 retains core themes and narrative structures, it systematically simplifies rhetorical devices, syntactic patterns, and metaphorical richness. The analysis contributes to debates on AI's role in literature by proposing a new typology of algorithmic adaptation ranging from faithful reproduction to generative transformation that extends existing frameworks in adaptation theory. This research demonstrates that GPT-4's stylistic simplification is not a correctable bug but a fundamental feature of its operation as a cultural agent, with profound implications for how literary heritage will be mediated in the digital age.