Selected Papers of #AoIR2024: The 25th Annual Conference of the Association of Internet Researchers Sheffield, UK / 30 Oct - 2 Nov 2024 Suggested Citation (APA): Kant, T. (2024, October). CheatGPT? The Realities of Automated Authorship in the UK PR and Communications Industries. Paper presented at AoIR2024: The 25th Annual Conference of the Association of Internet Researchers. Sheffield, UK: AoIR. Retrieved from http://spir.aoir.org. CHEATGPT? THE REALITIES OF AUTOMATED AUTHORSHIP IN THE UK PR AND COMMUNICATIONS INDUSTRIES Tanya Kant University of Sussex Introduction ChatGPT and its competitors have caused flurries of speculation in the creative industries for their ability to generate ‘human-like’ written text in the form of press releases, articles and other online content. However, like many apparently game- changing tools before them, current speculations about what generative text AI can do seem to be outweighing actual uses. This paper looks to refine and extend research on ethical AI use, authorship and generative text tools to ask: how do (human) content writers navigate and utilise generative text AI in current communications climates? How are employers of content writers negotiating the issues that ChatGPT and its competitors appear to bring? And what socio-cultural implications for authorship, creativity and ethical AI use do such tools create? Drawing on interview and survey data from content writers and their managers in the UK communications industries, I will present findings that explore sector-specific engagements with generative text AI. The project will consider frameworks of algorithmic and AI literacy (Cotter, 2020; Swart, 2021, Cox, 2024) to evaluate opportunities for empowering as-yet disempowered stakeholders of AI tools in creative industry sectors. Critical context The potential ‘dangers’ of AI have dominated popular and industry-based narratives in recent months, with global players framing AI tools as existential threats not just to human authorship but humanity itself (Roose, 2023). Despite the allure of these dominant narratives, scholars such as Leslie (2019) highlight that meaningfully assessing the effects of AI necessitates a move away from macrocosmic speculations and towards usage in specific contexts. This paper responds to such a need by exploring the realities of generative text AI usage in the UK communications and PR sector. Skujve et al. note that though there are numerous technical and theoretical http://spir.aoir.org/ studies of conversational generative AI, at present ‘there is a relative lack of knowledge as to why people use this technology’ (2023, n.p.). In the UK swathes of the communications sector are comprised of small- to medium-sized employers (herein SME) (TechUK, 2023). These stakeholders are reportedly increasing their usage of ChatGPT and its competitors (Kratchounova, 2023) but have little power in shaping the wider techno-political landscape of generative AI development. This is evidenced in the publication of the #AISummitOpenLetter created by numerous civil society organisations and campaigners (Connected By Data, The Trades Union Congress et. al, 2023), which highlights that in current AI development, ‘small businesses and artists are being squeezed out… as a handful of big tech companies capture even more power and influence’ (ibid, 2023). To critically analyse the realities of using ChatGPT I draw on frameworks of algorithmic and AI literacy. Dogruel et al define algorithmic literacy as ‘being aware of the use of algorithms in online applications, platforms, and services, knowing how algorithms work, being able to critically evaluate algorithmic decision-making as well as having the skills to cope with or even influence algorithmic operations’ (Dogruel et al., 2021, 4). At present work on algorithmic literacy is ‘rich but scattered’ and tends to focus on ‘awareness’ rather than ‘how to improve’ algorithmic literacy in specific contexts (Oeldorf-Hirsch & Neubaum, 2023, 16). There is emerging research that considers algorithmic literacy specifically in AI tool users – for example Cox uses the term ‘generative AI literacy’ (2023) to describe user awareness of ‘ethical and societal’ implications of generative AI – but to date there is little empirical research in this field.This project looks to map generative AI usage and understanding of smaller stakeholders in UK communications sectors in ways that might empower such stakeholders to use generative text AI tools ethically and critically. Methodology and findings The paper presents findings from a survey of 1,074 respondents - 637 content writers working in UK PR and communications roles or in the sector, and 437 managers of content writers. Interview responses are drawn from 21 interviews with selected survey respondentsi. Launched in collaboration with communications consultancy Magenta Associates, the project secured AHRC Impact Accelerator funding and has been awarded ethical approval via the researcher’s institutional ethics board. The project finds that generative text AI use is widespread in the sector, with 80% of content writers using gen AI tools frequently. We find that generative text AI is increasingly being used by content writers in ways that challenge speculative forecasts; these tools are useful for saving time, elements of idea generation and synthesising existing text, but are for the most part not (yet) used to generate press-ready, authorially convincing and stylistically branded outputs. We find that contrary to press reportage, most workers disclosure using generative text AI tools to others in their workplace. Qualitative data suggests that those who are using tools in total secrecy (8%) are doing so out of ‘embarrassment’ rather than to gain competitive advantage over co-workers. Finally, though we find the majority respondents believe generative text AI use is ethical, many writers and managers lack the assurances needed to confidently negotiate the legal, IP and employment implications of generative AI tools. I argue that ‘critical algorithmic literacy’ (Cotter, 2020) is best placed to empower SME https://www.forbes.com/sites/forbesagencycouncil/people/lorakratchounova1/ https://journals-sagepub-com.sussex.idm.oclc.org/doi/full/10.1177/14614448231182662#bibr22-14614448231182662 communications workplaces to create new avenues in AI understanding, authorship and practice, as this form of literacy best acknowledges unequal distribution of algorithmic power between different kinds of global and local stakeholders. In doing so I argue that AI-focused literacy is not just a question of upskilling or training but a question of challenging power relations. References Christian, A. (2023, October 25). The employees secretly using AI at work. BBC. https://www.bbc.com/worklife/article/20231017-the-employees-secretly-using-ai-at-work Connected By Data, The Trades Union Congress, Open Rights Group et.al. (2023, October 30) AI Safety Summit Open Letter to the UK Prime Minister. https://ai-summit- open-letter.info/ Cotter, K. M. (2020). Critical algorithmic literacy: power, epistemology and platforms: Michigan State University. Information and Media. Cox, A. (2024). Algorithmic Literacy, AI Literacy and Responsible Generative AI Literacy. Journal of Web Librarianship, 1–18. https://doi- org.sussex.idm.oclc.org/10.1080/19322909.2024.239534 Dogruel, L., Masur P., Joeckel S (2021). Development and validation of an algorithm literacy scale for Internet users. Communication Methods and Measures 16(2): 115– 133. Kratchounova, L. (2024, February 23). GenAI is changing public relations forever, but not how you may think. Forbes. https://www.forbes.com/sites/forbesagencycouncil/2024/02/26/genai-is-changing-public- relations-forever-but-not-how-you-may-think/ Leslie, D. (2019). Understanding artificial intelligence ethics and safety: A guide for the responsible design and implementation of AI systems in the public sector. The Alan Turing Institute. https://doi.org/10.5281/zenodo.3240529 PR Week staff. (2023, January 18). Will ChatGPT replace pros? PR Week. https://www.prweek.com/article/1810634/will-chatgpt-replace-pr-pros Oeldorf-Hirsch, A., & Neubaum, G. (2023). What do we know about algorithmic literacy? The status quo and a research agenda for a growing field. New Media & Society, 0(0). https://doi-org.sussex.idm.oclc.org/10.1177/14614448231182662 Roose, K. (2023, May 30). AI poses ‘risk of extinction’. New York Times. https://www.nytimes.com/2023/05/30/technology/ai-threat-warning.html Skjuve, M., Brandtzaeg, P. B., & Følstad, A. (2024). Why do people use ChatGPT? Exploring user motivations for generative conversational AI. First Monday, 29(1). https://doi.org/10.5210/fm.v29i1.13541 https://www.bbc.com/worklife/article/20231017-the-employees-secretly-using-ai-at-work https://ai-summit-open-letter.info/ https://ai-summit-open-letter.info/ https://doi-org.sussex.idm.oclc.org/10.1080/19322909.2024.239534 https://doi-org.sussex.idm.oclc.org/10.1080/19322909.2024.239534 https://www.forbes.com/sites/forbesagencycouncil/people/lorakratchounova1/ https://www.forbes.com/sites/forbesagencycouncil/2024/02/26/genai-is-changing-public-relations-forever-but-not-how-you-may-think/ https://www.forbes.com/sites/forbesagencycouncil/2024/02/26/genai-is-changing-public-relations-forever-but-not-how-you-may-think/ https://doi.org/10.5281/zenodo.3240529 https://www.prweek.com/article/1810634/will-chatgpt-replace-pr-pros https://doi-org.sussex.idm.oclc.org/10.1177/14614448231182662 https://www.nytimes.com/2023/05/30/technology/ai-threat-warning.html https://doi.org/10.5210/fm.v29i1.13541 Swart J., (2021). Experiencing algorithms: How young people understand, feel about, and engage with algorithmic news selection on social media. Social Media + Society, 7(2). https://doi-org.sussex.idm.oclc.org/10.1177/20563051211008828 i Project interviews are still on-going and so data here represents initial responses. All interviewee names have been changed to protect anonymity. https://doi-org.sussex.idm.oclc.org/10.1177/20563051211008828