Throughout
January 2024, the Society
of Authors (a UK union of professional literary writers,
illustrators and translators) conducted a survey among its 12,500
members and other authors, receiving almost 800 responses about respondents' experiences
with the generative artificial intelligence (AI) systems , and their
opinions and concerns about the future impact on creative careers.
The
results demonstrate not only the deep uncertainty about the future role of
generative AI in the profession, but also the impact it is already having on
careers and livelihoods. While some respondents are starting to use generative AI as a tool
in their work by choice, others – specifically, some translators and
illustrators – are being asked to use it by publishers and commissioning
organizations.
- About 1 in 5 respondents (22%)
say they have used generative AI in their work.
- This includes 1 in 10
illustrators (12%), a third of translators (37%), a fifth of fiction
writers (20%) and around a quarter of non-fiction writers (25% ).
- About 3 in 10 illustrators and
writers (31%) said they have used generative AI to provide ideas.
- About 1 in 10 translators (8%)
and a smaller proportion of illustrators (5%) said they have used
generative AI in their work because their editor or commissioning
organization asked them to.
Concern
about the impact of generative AI on creative careers includes groups of
authors who are already experiencing job loss , or the devaluation of their work, as a direct
result of new technologies.
- A quarter of illustrators (26%)
and more than a third of translators (36%) have already lost work due to
generative AI.
- More than a third of
illustrators (37%) and more than 4 in 10 translators (43%) say their job
income has decreased due to generative AI.
- Nearly two-thirds of fiction
writers (65%) and more than half of non-fiction writers (57%) believe that
generative AI will have a negative impact on future income from their
creative work, rising to more than three quarters in the case of
translators (77%) and illustrators (78%).
- More than 8 in 10 respondents
(86%) are concerned that their style, voice and image will be imitated or
reproduced in generative AI results.
- More than 8 in 10 respondents
(86%) are concerned that the use of generative AI will devalue creative
work done by humans.
- Some respondents are concerned
that generative AI could replace human creators, especially in areas such
as writing and content creation. They worry that this will lead to a
decline in quality and diversity in the creative industries.
- Even respondents most optimistic
about using generative AI systems ethically (for example, as tools to
improve efficiency and accessibility) reiterated that ethical concerns are
a major reason for avoiding the use of AI systems. generative in this
phase.
There is
almost unanimous consensus among respondents on the need to regulate generative
AI : ensuring
that consent is sought from copyright holders before their work is used to
develop systems, that credit and compensation are granted, and that the results
of generative AI systems are labeled as such.
- Almost all respondents (94%)
want to be given credit, compensated, and asked for their consent (95%)
when their work is used to develop generative AI systems or to enable the
production of generated results by AI.
- Almost all respondents (95%)
call for the Government to introduce safeguards and regulation to ensure
compliance with these consent, compensation and transparency measures.
- In a clear message to publishers
and other organizations about the importance of transparency in all uses
of AI software, the overwhelming majority of respondents (more
than 9 in 10) believe that publishers and other organizations should
clearly indicate highlighted when generative AI has been used to help with
audio, video, covers and illustrations, decision making, editing and
translation.
- Respondents also expressed
ethical concerns about generative AI systems, with many highlighting
biases and inaccuracies in AI-generated content, concerns about copyright
infringement, misuse of personal data, and exploitation of information.
works of other creators without consent or remuneration.
- Nearly all respondents (97%)
believe consumers deserve transparency and should know when generative AI
systems have generated all or part of what they read, see or hear.