Neither artificial or intelligent: Applications of creative machine learning in screenwriting and story development in the Charismatic consortium

Marshall, Kingsley ORCID: https://orcid.org/0000-0002-2518-7305, Jones, Rachael and Barley, Scott (2025) Neither artificial or intelligent: Applications of creative machine learning in screenwriting and story development in the Charismatic consortium. Journal of Screenwriting (Special Edition): Screenwriting and AI: Emerging Theories, Modes, and Practices, 16 (3). ISSN 1759-7137 (In Press)

Abstract / Summary

Introduction
“Until you were born, robots didn't dream, robots didn't desire unless we told them what to want.” – Professor Hobby (A.I. Artificial Intelligence, Spielberg, 2001).

The emergence of Fourth Industrial Revolution (4IR) technologies — from machine learning and AI to augmented reality, cyber-physical systems, and genomics — has triggered profound shifts in politics, economics, and creativity. Across all sectors, these technologies are reshaping workflows, business models, and cultural production. The screen industries are no exception. In recent years, 4IR tools have transformed how screen media is created, produced, and consumed, influencing everything from creative workflows to the growing dominance of technology companies including Netflix. These changes have prompted both industrial action and deep unease among creative communities.

In a sector already under pressure from declining advertising revenues and fast-changing distribution models, media companies now face the imperative to “produce more, faster, and manage the growing complexity of audiences, channels, and technologies” (Pfeiffer Reports 2018: 2). A slew of technology companies and startups have stepped in offering solutions which promise efficiencies at all stages of the production chain - from development and screenwriting and, to production and post-production right through to distribution, marketing and exhibition. These often take the form of opaque, general-purpose ‘black box’ tools, concealing both the methods by which content is generated and the datasets it draws upon. In the high-risk stage of screen development, where intellectual property and chain of title are critical, such apacity has created a great deal of anxiety among writers and presents what appears to be an existential threat to those who commission their labour - agencies, studios, broadcasters, and streaming services – all operating in a confusing regulatory environment (Sweney 2025, Stacey and Courea 2025). At the time of writing, multiple legal cases are underway that test the copyright and copyright-adjacent implications of the datasets and algorithmic models used in AI (Mishcon de Reya 2025, BakerHostetler 2025).

In light of this, screenwriters, commissioners, and producers are rightly demonstrating caution as to what the use of these models means in terms of legal and regulatory compliance, and the use and retention of intellectual property in what has been described as a creative ‘Wild West’ (Stelter 2024).
At the same time, some see in these technologies the potential for new forms of storytelling, for new audiences: “Far from heralding the end of human creativity, AI presents new ways of being original” (Mökander et al., 2025).

This paper draws on the early findings of the £1.04m Innovate UK-funded Charismatic Consortium, a year-long research and development programme exploring the design, deployment, and ethical use of AI tools for screenwriting and story development. Our research examines the design, deployment, and effectiveness of new and established AI tools specifically relating to screenwriting and story development, and the potential ethical application of wider 4IR technologies within the screen industries. Our research examines the gap between what screenwriters and filmmakers aspire to create and what can realistically be codified in AI systems. It also identifies the frictions between the vocabulary and practices of screen creatives and those of software engineers, and considers how to bridge them. Central to our aim was giving voice to creatives in a technology landscape dominated by companies whose rapid product releases often disregard — and sometimes actively seek to dismantle — long-established rights and protections for creative content.

While AI may increasingly enable aspects of creative practice, the ability to articulate why a creative work was made, and to transparently demonstrate how it was made, remains as important as ever to writers and those who commission them as the outputs themselves. From the perspective of working filmmakers, we consider the impact of these technologies on screen story development and delivery, and their broader implications for the future of creative authorship.

Item Type:	Article
Additional Information:	© Kingsley Marshall 2025. The definitive, peer reviewed and edited version of this article is published in Journal of Screenwriting, volume 16, issue 3, pages, year, DOI link.
Uncontrolled Keywords:	artificial intelligence, AI, screenwriting, film, television, writing, development innovate uk, machine learning, production cultures, production, charismatic, charismaAI, channel 4, aardman
ISSN:	1759-7137
eISSN:	1759-7145
Subjects:	Film & Television > Television
Department:	School of Film & Television
Depositing User:	Kingsley Marshall
Date Deposited:	22 Sep 2025 11:27
Last Modified:	22 Sep 2025 11:27
URI:	https://repository.falmouth.ac.uk/id/eprint/6156

View Record (staff only)