Across the UK, people are navigating AI from a position of both opportunity and anxiety. The BFI's 2025 report AI in the Screen Sector: Perspectives and Paths Forward called for ethical and commercially viable frameworks for AI supported production, including proper licensing of creative IP used to train AI models. Unions including Equity fight to ensure their members are protected from AI misuse and monitor how AI is deployed in relation to their creative identities and rights. Most businesses expect AI to change some working practices, yet many remain uncertain about how adoption will unfold. Human oversight and interpersonal skills are often identified as the most important assets for adapting to these changes.

The Film & TV Charity's Principles for Mentally Healthy Productions asks that demands on workers are managed in ways that protect wellbeing and performance. AI tools that are poorly implemented, e.g. by creating ambiguity about roles, generating additional checking and correction burdens, or introducing constant context-switching, can quietly increase cognitive load rather than reduce it.

So, before adopting any AI tool, businesses should ask whether this genuinely reduces pressure on our team, or does it just shift effort into less visible places?

Team structure and role clarity is also important. When AI is introduced into a production workflow, the question of who is responsible for its outputs, who has the authority to reject them, and whose creative judgement takes precedence must be addressed. Ambiguity about AI assisted tasks can erode the psychological safety that supports good creative work.

CIISA Standards are clear that your team must be free from exposure to psychological harm which is a standard that applies to management practices enabled by AI, not just interpersonal behaviour.

AI tools that enable surveillance, monitoring of productivity, or automated performance assessment can create environments of psychological harm if they are deployed without transparency, sensitivity and consent.

AI trained on unrepresentative data can embed and amplify biases in casting suggestions, script analysis, hiring tools or content recommendations. An inclusive working environment requires active human steps to prevent discrimination. 

If an AI tool contributes to a problematic outcome such as a biased recommendation, some inappropriate content generation, or even a failure of duty of care, everyone needs to know they can raise this through existing reporting structures and that it will be taken seriously.

Dignity and respect must extend to all working relationships and to all the tools and systems that shape those relationships. So, AI systems used in hiring, scheduling, feedback or performance review carry the same duty of care as any manager or supervisor.
 

AI will continue to change how we work. We must shape that change around the values of dignity, safety, fairness and creative respect that the screen sector has worked hard to articulate. That requires the same thing all good working practice requires; leadership, transparency, inclusion and people who feel safe enough to speak up when something isn't right.