ACE urges caution on using generative AI in funding bids

Organisations applying for funding from Arts Council England (ACE) should use generative AI “cautiously”, the funding body has warned.

The arts council published a statement this week outlining its position on the use of AI in writing funding bids and in processing, assessing and making decisions on applications.

The statement said the arts council recognised that “some applicants may choose to use generative AI to support the drafting of an application that expresses their own, original idea, and some grantholders may use it to assist with the organisation of their reporting material”, and said it understood that AI technologies could be useful in “streamlining reports, and/or the process of applying”.

However the arm’s-length body urged those using such tools to “do so cautiously”, emphasising that applicants and grantholders are accountable for what they submit.

ACE said it was increasingly aware that the use of generative AI tools can lead to submissions “with similar language and text”, which it said can “distract from an application’s unique proposition”.

“We advise applicants to consider whether an application drafted with AI support offers the best and clearest representation of their project, and to ensure that any application submitted to ACE represents a true, original, and deliverable proposition,” said the statement.

The arts council also warned that generative AI tools “create potential risks around bias, transparency, data protection and the moral and legal rights of creators”.

It urged applicants and grantholders to employ such tools “in ways that align with existing data protection policies, and with their own, or their organisation’s, values”.

The statement advised applicants to be cautious of using systems that retain users’ information for the purpose of training commercial technology companies’ models.

The arts council said that decisions about funding applications would always be made by people to ensure the process is fair, equitable and transparent.

The statement said: “Assessment of Arts Council England funding applications is carried out by arts council staff, who are experts in their field. We do not use AI technologies in carrying out the assessment, or decision-making on applications.”

However, the statement said that the arts council may consider using AI to manage the “operational processing of applications”.

“We may explore using AI over the coming months to help us to carry out some processing tasks more efficiently and effectively,” said the statement. “This could include carrying out administrative tasks such as checking and verifying information, and analysing and categorising data sets by, for example, location or demographics.”

The position statement comes in response to research being undertaken by ACE into the use of AI in the culture sector, which found there was a “pressing need” for clear guidance around the use of AI in the arm’s-length body’s funding processes.

The arts council’s director for new technologies and innovation, Owen Hopkin, said the questions raised by AI were “critical” for those working in the culture sector.

In a blog outlining ACE’s approach, Hopkin wrote: “How can we protect the intellectual property of creative practitioners? What do these new tools mean for bias and authorship? What is the environmental impact of the energy they need to power their work? And – just as importantly – what potential do they offer for accessibility, productivity and creativity?

“We know that many in the cultural sector are already beginning to explore these questions, and that organisations across the sector are thinking seriously about both the promise and the challenges that AI presents for them.”

ACE has been exploring these questions over the past 12 months with support from researcher and senior Goldsmiths lecturer Oonagh Murphy, who has joined the arts council team as a fellow from the Arts and Humanities Research Council-sponsored initiative, Bridging Responsible AI Divides (Braid).

This work has “supported us to consult with colleagues, peers in other public bodies, and partners in the cultural sector, to test our thinking on the shared challenges and opportunities we were all facing”, said Hopkin. 

The question of copyright and intellectual property had come through as one of the most pressing issues in the wider AI debate, Hopkin said.

“The artists and creators with whom we’ve consulted have been clear about their urgent concern around how their work may be ingested, or replicated, by large language models. As part of the UK Government’s recent consultation on AI and intellectual property we shared insights from our own conversations with the sector about the need to protect copyright and authorship – both legally, and as the foundation of our country’s creative and cultural life.”

The funding body has also published its first internal staff policy on the responsible use of AI.

“While our public statement addresses external questions, this internal-facing policy was created to help staff navigate the day-to-day use of AI tools at work,” said Hopkin. “It highlights the risks and responsibilities associated with these technologies – from data protection to environmental impact – and encourages informed use of such tools in appropriate areas of work.”

He added: “It is not about mandating use or avoiding it. It is about supporting staff across all areas of the organisation to use these tools in the right context, carefully, ethically and transparently if and when they choose to do so.” 

Hopkin said the arts council’s goal was to be “receptive, responsive and responsible as AI becomes a greater part of our lives and the sector’s experience grows with it”.

“These documents are not intended to be a final word. We’ll continue to review and refine our work in this area while welcoming the views of those we work with,” he said. “We will also look to roll out resources and provide assistance and advice where we can. The technology is evolving and our approach will need to evolve too.”