Clementine Collett, BRAID (Bridging Responsible AI Divides) Fellow at the Minderoo Centre for Technology and Democracy, University of Cambridge, wrote a November 25, 2025 essay for The Conversation discussing the impulse that led to her latest research on AI, Note: Links have been removed,
Back in 2023, I was completing my doctorate on AI and gender bias and my debut novel, Something About Her, had just been published. It was also the year that many prominent authors including Jodi Picoult, John Grisham and George R.R. Martin filed a lawsuit against OpenAI for using their work to train generative artificial intelligence (a type of AI that creates new content based on user prompts) without permission. This case is still proceeding through the courts, as are many others on similar grounds.
At the time, I remember thinking: we desperately need to know more about the implications of generative artificial intelligence (GenAI) for the novel. As a writer, avid reader and an academic researching the impact of AI on work and society, this felt like a pivotal and crucial time to investigate this.
Particularly as governments around the world continue to ask big questions about how AI is regulated, designed and used.
In 2024, I started researching how hundreds of novelists and publishers were using GenAI across the UK. This consisted of a major survey with 332 literary creatives (258 published novelists, 42 professionals in fiction publishing, 32 literary agents for fiction); six focus groups with 52 writers and publishers; interviews and case studies with creatives and industry experts; and a forum which brought different stakeholders together to discuss the changes and challenges which GenAI is posing to the novel.
What I found was starker than I had anticipated.
The findings, published in my report, The Impact of Generative AI on the Novel, show just how urgently people in the industry want guardrails to be put around AI to help protect this precious art form.
…
I have more about Collett’s research from a University of Cambridge November 20, 2025 press release (also on EurekAlert but published on November 19, 2025) by Fred Lewsey, Note: Links have been removed,
A new report involving hundreds of literary creatives from across the UK fiction publishing industry reveals widespread fears over copyright violation, lost income, and the future of the art form, as generative AI tools and LLM-authored books flood the market.
Just over half (51%) of published novelists in the UK believe that artificial intelligence is likely to end up entirely replacing their work as fiction writers, a new University of Cambridge report shows.
Close to two-thirds (59%) of novelists say they know their work has been used to train AI large language models (LLMs) without permission or payment.
Over a third (39%) of novelists say their income has already taken a hit from generative AI, for example due to loss of other work that facilitates novel writing. Most (85%) novelists expect their future income to be driven down by AI.
In new research for Cambridge University’s Minderoo Centre for Technology and Democracy (MCTD), Dr Clementine Collett surveyed 258 published novelists earlier this year, as well as 74 industry insiders – from commissioning editors to literary agents – to gauge how AI is viewed and used in the world of British fiction. *
Genre authors are considered the most vulnerable to displacement by AI, according to the report, with two-thirds (66%) of all those surveyed listing romance authors as “extremely threatened”, followed closely by writers of thrillers (61%) and crime (60%).
Despite this, overall sentiment in UK fiction is not anti-AI, with 80% of respondents agreeing that AI offers benefits to parts of society. In fact, a third of novelists (33%) use AI in their writing process, mainly for “non-creative” tasks such as information searches.
However, the report outlines profound concerns from the cornerstone of a publishing industry that contributes an annual £11bn to the UK economy, and exports more books than any other country in the world.
Literary creatives feel that copyright laws have not been respected or enforced since the emergence of generative AI. They call for informed consent and fair remuneration for the use of their work, along with transparency from big tech companies, and support in getting it from the UK government.
Many warn of a potential loss of originality in fiction, as well as a fraying of trust between writers and readers if AI use is not disclosed. Some novelists worry that suspicions of AI use could damage their reputation.
“There is widespread concern from novelists that generative AI trained on vast amounts of fiction will undermine the value of writing and compete with human novelists,” said Dr Clementine Collett, BRAID UK Research Fellow at Cambridge’s MCTD and author of the report, published in partnership with the Institute for the Future of Work.
“Many novelists felt uncertain there will be an appetite for complex, long-form writing in years to come.”
“Novels contribute more than we can imagine to our society, culture, and to the lives of individuals. Novels are a core part of the creative industries, and the basis for countless films, television shows, and videogames,” said Collett.
“The novel is a precious and vital form of creativity that is worth fighting for.”
Tech companies have the fiction market firmly in their sights. Generative AI tools such as Sudowrite and Novelcrafter can be used to brainstorm and edit novels, while Qyx AI Book Creator or Squibler can be used to draft full-length books. Platforms such as Spines use AI to assist with publishing processes from cover designs to distribution.
“The brutal irony is that the generative AI tools affecting novelists are likely trained on millions of pirated novels scraped from shadow libraries without the consent or remuneration of authors,” said Collett.
Along with surveying a total of 332 literary creatives, who participated on condition of anonymity, Collett conducted focus groups and interviews around the country, and convened a forum in Cambridge with novelists and publishers. **
Many novelists reported lost income due to AI. Some feel the market is increasingly flooded with AI-generated books, with which they are forced to compete. Others say they have found books under their name on Amazon which they haven’t written.
Some novelists also spoke of online reviews with telltale signs of AI, such as jumbled names and characters, that give their books bad ratings and jeopardise future sales.
“Most authors do not earn enough from novels alone and rely on income streams such as freelance copywriting or translation which are rapidly drying up due to generative AI,” said Collett. ***
Some literary creatives envisioned a dystopic two-tier market emerging, where the human-written novel becomes a “luxury item” while mass-produced AI-generated fiction is cheap or free.
When it came to working practices, some in the study consider AI valuable in speeding up repetitive or routine tasks, but it was seen to have little to no role to play in creativity.
Almost all (97%) novelists were “extremely negative” about AI writing whole novels, or even short sections (87% extremely negative). The aspects novelists felt least negative about using AI for were sourcing general facts or information (30% extremely negative), with around 20% of novelists saying they use AI for this purpose.
Around 8% of novelists said they use AI for editing text written without AI. However, many find editing to be a deeply creative process, and would never want AI involved. Almost half (43%) of novelists felt “extremely negative” about using AI for editing text.
Forum participant Kevin Duffy, founder of Bluemoose Books, the publisher behind novels such as The Gallows Pole and Leonard and Hungry Paul – both now major BBC TV dramas – is on record in the report saying:
“[W]e are an AI free publisher, and we will have a stamp on the cover. And then up to the public to decide whether they want to buy that book or not. But let’s tell the public what AI is doing.” Many respondents echoed this sentiment. ****
The research found widespread backlash against a “rights reservation” copyright model as proposed by the UK government last year, which would let AI firms mine text unless authors explicitly opted out.
Some 83% of all respondents say this would be negative for the publishing industry, and 93% of novelists said they would ‘probably’ or ‘definitely’ opt out of their work being used to train AI models if an opt out model was implemented.
The vast majority (86%) of all literary creatives preferred an ‘opt in’ principle: rights-holders grant permission before AI scrapes any work and are paid accordingly. The most popular option was for AI licensing to be handled collectively by an industry body – a writers’ union or society – with half of novelists (48%) selecting this approach.
“Our creative industries are not expendable collateral damage in the race to develop AI. They are national treasures worth defending. This report shows us how,” said Prof Gina Neff, Executive Director of the Minderoo Centre for Technology and Democracy.
Some novelists worry AI will disrupt the “magic” of the creative process. Stephen May, writer of acclaimed historical novels such as Sell Us the Rope expressed anxiety over AI taking the required “friction” and “pain” out of a first draft, diminishing the final product.
“Novelists, publishers, and agents alike said the core purpose of the novel is to explore and convey human complexity,” said Collett. “Many spoke about increased use of AI putting this at risk, as AI cannot understand what it means to be human.”
Authors fear AI may weaken the deep human connection between writers and readers at a time when reading is already at historically low levels, particularly among the next generation: only a third of UK children say they enjoy reading in their free time.
Many novelists want to see more AI-free creative writing on the school curriculum, and government-backed initiatives aimed at finding new voices from underrepresented groups to counter risks of “homogeneity” in fiction brought about by generative AI.
The research reveals a sector-wide belief that AI could lead to ever blander, more formulaic fiction that exacerbates stereotypes, as the models regurgitate from centuries of previous text. Some suggest the AI era may see a boom in “experimental” fiction as writers work to prove they are human, and push the artistry further than AI.
“Novelists are clearly calling for policy and regulation that forces AI companies to be transparent about training data, as this would help with the enforcement of copyright law,” added Collett.
“Copyright law must continue to be reviewed and might need reform to further protect creatives. It is only fair that writers are asked permission and paid for use of their work.”
Notes:
* Of the published novelists involved in the survey, 90% are published “traditionally” through publishing houses, while 10% are self-published.
** This consisted of 32 literary agents for fiction (9%), 258 published novelists (78%), and 42 professionals who work in fiction publishing (13%). The survey took place between February and May 2025.
*** The median income for an author in the UK in 2022 was just £7,000, far lower than the minimum wage. [Thomas, A., Battisti, M., & Kretschmer, M. (2022). UK Authors’ Earnings and Contracts 2022.]
**** Full quote from Kevin Duffy as featured in the report: “’For a small independent publisher of literary fiction like Bluemoose Books, our only stand is to say we don’t want any part of this, we are AI free, and we are an AI free publisher, and we will have a stamp on the cover. And then up to the public to decide whether they want to buy that book or not. But let’s tell the public what AI is doing. It’s got brilliant capacity to do fantastic things in other avenues, but for the creative industries and for literary fiction in particular, it is very limited.”
The full report, The Impact of Generative AI on the Novel, will be published on the Minderoo Centre for Technology and Democracy website: https://www.mctd.ac.uk/
The research in the report was supported by the BRAID (Bridging Responsible AI Divides) programme with funds received from the Arts and Humanities Research Council [grant number AH/X007146/1]. This report is published in association with the Minderoo Centre for Technology and Democracy at the University of Cambridge, and with the Institute for the Future of Work.
The Cambridge University’s Minderoo Centre for Technology and Democracy (MCTD) November 20, 2026 report “Impact of Generative AI on the Novel” by Dr. Clementine Collett has a dramatic summary, Note: Links have been removed,
The rise of generative AI has positioned the UK publishing sector on the brink of irreversible change
The UK’s creative industries are globally renowned and world leading, contributing £126bn gross value added to the UK economy annually and an abundance of soft power.
Novels in particular play an enormous role in the creative industries and the publishing sector. Sales and exports of novels are significant, and they also provide the basis for countless other outputs, including films, television, music, plays, videogames, and more. They contribute enormously to our culture, wellbeing, personal development, and entertainment. The risk of harm or even displacement of the novel, therefore, should not be underestimated.
Yet, so far, UK Government policy has focused largely on the economic growth that GenAI might unlock for the economy through innovation and increased productivity, without full consideration of how this could detrimentally impact the creative industries, including novelists, the publishing sector, readers, and the novel itself.
In this groundbreaking new report, made possible by funding from BRAID [Bridging Responsible AI Divides)] UK and published in association with the Minderoo Centre for Technology and Democracy and the Institute for the Future of Work, Dr Clementine Collett examines how industry practitioners, from novelists to publishers, are experiencing AI’s significant impacts. She looks at how literary creatives are using (or not using) GenAI, how their work and income are being affected, how this is shaping their hopes and fears for the future, and what measures they would like to see implemented by the Government to protect the UK’s precious literary scene.
Key Findings
Use of Gen AI
Most literary creatives are not using GenAI. 67% of novelists said they never use AI, along with 66% of literary agents, and 55% of those working in fiction publishing. This is due to the moral and environmental implications, the negative perception of using AI for ‘creative’ tasks, and concerns around the inaccuracies of GenAI. Some creatives did not take issue with the use of AI per se but took issue with systems which had been trained on unlicensed data.
Some literary creatives use GenAI for ‘non-creative’ tasks. 33% of novelists report that they use GenAI within their work, along with 34% of literary agents, and 45% of those working in fiction publishing. Most commonly, these literary creatives use GenAI for ‘non-creative’ tasks such as information search.
Social Concerns
Literary creatives warned of the loss of creativity and the de-skilling of younger generations. There was concern that increased use of GenAI would lead to a loss of creative practice and de-skilling in terms of creativity, imagination, empathy, resilience, problem-solving, and critical thinking. Much of this was discussed in reference to younger generations.
Destruction of trust and connection was a major concern amongst literary creatives. Opacity and uncertainty around how writers are using GenAI was thought to endanger the connection between writers and readers. Moreover, this was predicted to sow mistrust within the industry. Novelists were concerned about false accusations around AI use damaging their reputation.
GenAI was seen to provide opportunity for those with accessibility needs, but also to perpetuate problematic stereotypes in its output. Participants perceived GenAI to provide important opportunities for voices in literature which might not otherwise be heard. However, there was also concern regarding the stereotypes and discrimination that output from these systems might perpetuate due to biased training data.
Literary Changes
Literary creatives predicted a loss of originality in literature. It was predicted that there would be a loss of originality in the content, style, and language of novels due to an increased use of GenAI in writing and publishing. However, there was also discussion about a potential rise in experimental fiction to counteract ‘AI style’.
Literary creatives predicted that genre fiction was at higher risk of displacement than literary fiction. In the survey, two thirds (66%) of literary creatives saw romance to be extremely threatened, closely followed by thrillers (61%), and crime (60%). 32% saw literary fiction to be extremely threatened.
Economic Impacts
Over a third (39%) of novelists reported that their income has already been negatively impacted by GenAI. This was mostly attributed to competition from AI-generated books and loss of jobs which provide supplementary streams of income.
Novelists fear their work will be displaced and their future ability to earn from writing will be entirely diminished. 85% of published novelists said they think their future income will be negatively impacted by GenAI, and over half (51%) of published novelists agreed that AI is likely to displace their work entirely.
Copyright and Data Use
Novelists prioritise control of their work, meaning asking permission would likely result in more access to data than a rights reservation model. 93% of novelists said they would ‘probably’ or ‘definitely’ opt out of their work being used to train AI models in contrast to 75% of novelists who said they would ‘probably not’ or ‘definitely not’ opt in. This shows that novelists prioritise control of their work and that asking creatives for permission to use their work would likely lead to more access to data for AI companies.
Most novelists (59%) report that their work has been used to train GenAI models without permission or remuneration. Almost two thirds of novelists (59%) reported that they know their work has been used to train AI models. Of these, 99% reported they did not give permission for this use and 100% reported they were not remunerated.
There is widespread backlash against a ‘rights reservation’ (opt-out) model. 83% of all literary creatives said that a data mining exception which allows rights holders to reserve their rights (an opt-out model) would be negative. This was mostly put down to underestimated negative impacts, technical infeasibility, and overblown expectations of growth.
Literary creatives consistently called for a licensing market which would enable them to give permission, have control, and be remunerated. 86% of literary creatives indicated that a model based on an ‘opt-in’ principle would be their preferred option. Almost half of novelists (48%) said they would want licensing to be negotiated collectively through a writers’ union or society. This was the most popular option in terms of how a licence would be negotiated.
Policy Recommendations
Copyright and Responsible Generative AI Design
1. Foster a Licensing Market and Continue to Review Sufficiency of Copyright Law:
The UK Government should not implement exceptions to copyright law or an opt-out model for AI training. Instead, the UK Government should reinforce copyright law through fostering a dynamic, accessible, and fair licensing market for AI training. Given recent case law, the Government should also continue to review whether copyright law is fit for purpose in the age of AI, and should potentially implement legislative reform to clarify the law and further protect creatives.
2. Mandate Granular Transparency and Fair Remuneration:
The UK Government should issue policy and regulation which requires granular transparency from technology companies on details of their training data for GenAI systems, and which requires the fair remuneration of creatives for the use of their work in training and fine-tuning GenAI models. This would help to facilitate a licensing market and enable creatives to exercise their rights.
3. Invest in Responsible SLM Development:
The UK Government should develop its Industrial Strategy, offer funding, and regulate data and technology to encourage investment in UK-founded small language models (SLMs) which are responsibly designed and trained. This should enable sophisticated models that are less damaging to the environment than LLMs. Moreover, they should be designed to ensure no discriminatory output. Social impacts should be ascertained, understood, and mitigated on an ongoing basis.
Education, Research and Funding
4. Include AI-Free Creative Writing and AI Critical Skills in School Curricula:
The Department for Education, and equivalent devolved bodies, should mandate that creative writing and reading is included in school curricula and that it does not include the use of GenAI (where access needs do not require its use). Alongside digital literacy skills, children should be taught skills which enable them to be critical of AI systems, so that they are informed about the limitations and downsides of GenAI use.
5. Fund Creative Writing Initiatives and Independent Publishers:
It is even more important in the GenAI era that the UK Government invests in creative writing initiatives and the work of cutting-edge independent publishers through increased funding to Arts Councils (England, Northern Ireland, Wales) and Creative Scotland. These funds should particularly focus on initiatives that encourage writing with vulnerable, minority, and underrepresented groups to counter specific risks around homogeneity of voices and disparity of resources which could be worsened by GenAI. This increased funding would help to ensure that unique and diverse voices in literature are encouraged and nurtured, and skills and craft are shared and passed on.
6. Increase Research on Market and Impacts:
UK Research and Innovation (UKRI) and the UK Government should increase support and funding for research on readers and creatives to aid the UK Government’s impact assessment on AI and copyright, and future decisions on AI and data regulation. Particularly, research should focus on reader preferences around AI content, the impact of GenAI on children’s creative development, and mapping the impacts of GenAI on creatives and on society more widely.
Publishing Industry Practice
7. Increase Industry Transparency:
Publishers should be transparent about their in-house use of AI. This should be openly communicated and discussed with literary agents and authors to increase trust, collaboration, and knowledge-sharing within the industry, and to allow for traceability as needed.
8. License GenAI Training Rights as Sub-Rights if going through Publishers:
Publishers should recognise and respect that GenAI training rights lie with the author and should license accordingly. The industry should work together to normalise that, if going through publishers, GenAI training rights should be licensed as sub-rights with a fair financial split.
9. Support Around AI Use and Contracts:
The industry should increase support and guidance around navigating the implications of AI use and AI training rights in contracts, particularly for self-published and un-agented authors, along with independent publishers. This support could include campaigns to encourage authors and publishers to join membership societies and trade unions. These bodies should continue guidance for their members on AI use, copyright, and contractual clauses.
If implemented, these recommendations would help the UK to continue to be a global superpower in the creative industries and a world leader in the design and implementation of responsible AI.
The novel is more than an art form. It has been the basis for a thriving British creative sector. It contributes more than we can imagine to our society, culture, and to the lives of individuals. It is a form of creativity which is precious and one which is worth fighting for.
…
Go to “The Impact of Generative AI on the Novel” and scroll down to the end of the page to access the full report.
