Library · AI in production

AI tools in production, assessed rather than admired.

This is the hardest list in the library to keep straight, because the subject is saturated with people selling things. A large proportion of free AI filmmaking content is an advert for a subscription, a demo of a capability that does not survive contact with a delivery specification, or a confident claim about the law by somebody who has not read it.

So this page is shorter than the others and weighted towards research and industry reporting rather than tutorials. It is for crew who want to know what is landing in production pipelines, and for anybody being told by somebody with a large following that their department is about to disappear.

Where to start

Read the sector research before you watch a tool demo.

01

Start with the UK sector research

The BFI has published substantial free research on AI in the screen sector, including recommendations covering rights, ethics, employment and environmental impact. It is the most sober thing available and it is free.

02

Then the workflow reality

The tools embedded in professional pipelines today are mostly unglamorous: rotoscoping and matting assistance, upscaling, denoise, speech cleanup, transcription and search. Learn those before the generative headlines.

03

Then decide what you will not do

The most useful thing you can develop here is a position: what you will use, what you will disclose, and what you will refuse. That is a professional judgement, and it is worth forming deliberately rather than by drift.

Futures in Film is openly pro-AI: we adopt, test, teach and build with these tools, and we publish our own position, including the refusals.

The list

Free, checked, deliberately short, and every one of these links out.

A large number of popular channels are not here because their teaching could not be told apart from their marketing. That is a statement about what we will put our name to rather than a claim that they are bad.

The national film body’s own report on AI in the UK screen sector, with recommendations on rights, ethics and environmental impact. Read the research before you watch a tool demonstration.

A joint report on where generative tools are actually being used and what that means for the people doing the work.

CoSTAR Network (opens in a new tab) · Free to read · UK research network

Public insight and lab outputs on convergent screen technologies, virtual production and real-time workflows. UK context, in a conversation otherwise conducted entirely in American.

Corridor Crew (opens in a new tab) · Free · YouTube · what is actually achievable

VFX artists breaking down effects work, including their own mistakes and including machine-assisted work. Clear about what a shot cost and how long it took, which is the question that matters.

Bectu (opens in a new tab) · Free to read · paid membership · anyone being asked to use these tools

The union’s position and guidance on AI at work, consent and rights. Read what your union says before your employer tells you what it thinks.

Writers’ Guild of Great Britain (opens in a new tab) · Free to read · paid membership · writers

Guild positions on AI, authorship and minimum terms. The published position is the benchmark you negotiate against.

Every link here goes to the owner’s own page, in a new tab. How we treat other people’s work · Last checked September 2026.

Money on it

Very little, and we would wait

This is the one subject where we would actively counsel patience about spending money. The tools and the pricing change faster than any course can be revised.

Tool subscriptions, once a job requires one

Worth paying for at the point a production asks you to deliver using a specific tool, and rarely before. Subscriptions bought speculatively during a hype cycle are the most reliably wasted money in this field.

Paid

Legal advice on rights and training data

Worth it for anybody producing commercially. The rights position around training data, likeness and generated assets is unsettled, and a free video is not the place to resolve it for your project.

Paid

Where the gap gets filled

By testing it in a real pipeline and writing down what happened.

Our AI and future workflows strand is built on testing rather than assertion: take a claim about a tool, put it through a real production task, and publish what happened, including the failures. That is slower and less exciting than a tutorial, and it is the only answer we are willing to give.

Everything we publish about this is free to read whether or not you ever train with us.

BSC Expo 2023 — interview with Rizwan Wadan

BSC Expo 2023 — interview with Rizwan Wadan

To be embedded once uploaded.