Should Filmmakers Use AI For Their Music Needs? Embracing A Changing World.
- James Smith
- Apr 14
- 4 min read
There are many articles outlining the excitement and pitfalls of generative AI. It is an uncertain time but in any situation AI is here and it is here to stay, so we can only embrace change and move with the times or fall by the wayside. We must, however, be aware of the ramifications of choosing to use AI or a human in the creative process.
With the ease of creating a large amount of music quickly at very low cost it may seem like a no-brainer to throw a few prompts into an engine and have it churn out a solution to your musical needs. So why wouldn’t filmmakers just use AI?
Authenticity
The primary motivating factor for any filmmaker is, or should be, quality and authenticity. With generative AI having a fundamental requirement to scrape and learn from existing material you will never get a unique score for your film even if the quality is deemed passable.
There is an argument where AI music can serve a purpose. For new starters and students who want to focus on honing single elements of their craft in a ‘finished’ perspective, AI music can provide that quick reference that puts a scene into general context.
A film is only as good as its weakest element so this can only be used to an extent before the film is held back by the music sounding bland and generic.
The Illusion of Competency
Knowing your skills and competency is always a great thing but the use of generative AI allows people with little to no knowledge to use AI software and come out with something that on the surface sounds convincing. It would be all too easy to fall into the trap of thinking that by using AI you have a new skill when in fact the feeling of being a music producer, graphic designer, storyteller is merely an illusion. Furthermore, the emulation generated has no context for its implementation so cannot respond to the story being told or direct nuances in subtext.
The difficulty lies in the juxtaposition between the ease of creation against the experience of how and when to use that sound. If you do not have a musical understanding the impulse will be to use AI for the easiest fix, however, this leaves the question ‘how do you know if the AI generated content is good enough for the scene?’.
What I am not saying is that everyone who uses AI is incompetent, but wish caution of the illusion that using AI creates. At some point in the process AI needs interaction and that needs a real person with an understanding of the art form.
Legality
Currently AI software is having a huge developmental boom and as such laws and legislation are fighting to keep up. The main focus that comes up is the grey area of who owns the music? Because grey areas are dangerous places in the creative industries it is no wonder production companies do not want to risk being involved with AI when it may have major legal ramifications down the line. For anyone wanting to take working in film seriously using generative AI tools is a one stop shop for all manner of problems and is a blanket no-go area. At least until proper legislation is put in place.
Within the music industry legal cases are rife with the theft of copyrighted material being used to feed AI software. For new creators it presents difficulties in not having opportunities to work on early years skills, and for the film music industry as a whole it adds a huge economic impact. I myself have seen a noticeable dip in royalty revenue since AI music has entered the market.
PRS for Music Economic study: https://www.prsformusic.com/m-magazine/news/cisac-generative-ai-study-music-creators-future-at-risk
Morality
Morally speaking, using generative AI in a communal creative process is fundamentally damaging to the team. Directors and producers are responsible for the creative output of the whole process and to take any one department of the film and replace this with AI sets a dangerous precedent that devalues that part of the creative process, the people and the film itself; shifting the value from creatives to tech companies.
At this moment it is more prevalent than ever and not only the music industry is at risk. The recent Seedance 2.0 Brad Pitt and Tom Cruise AI generation not only generated elements but stole the entity of whole people
Financials
At some point in the film industry things boil down to the budget. Every film has one and someone is responsible for balancing the books. The decision to hire a music team or use AI is now a regular consideration by many filmmakers, what with reduced production costs, faster turn around times, lower licence fees.
This however, comes at a big risk. You can create easily and cheaply, the quality may even be a decent standard but the risk of copyright infringement is real. Currently there is a lack of copyright protection and unclear ownerships and the end user will be liable. Using AI and profiting from a film is a surefire way to be met with lawsuits or legal penalties if the music mimics any copyrighted work.
To summarise, the choice is down to the filmmaker to make decisions for the benefit of their film. Do so with open eyes and an awareness of the risks involved.
The use of AI as a tool can definitely be a useful and supportive aspect of emphasising the form of human expression, aiding the refinement of music, but the use of generative AI strips away the collaborative approach to storytelling.
This is fundamental to the efficacy of filmmaking and can only lead to limiting the overall validity of the production. Listen to some audio demos of AI by the Northern Film Orchestra: https://www.northernfilmorchestra.com/post/article-ai-in-film-music-should-composers-embrace-it-or-fear-it
Testing out Seedance2.0 - Trying this out we used our logo plus the super basic prompt ‘turn the logo into a mini film about AI shooting a movie’
