Every week there’s a new AI tool promising to revolutionise film post-production. Most of them are overhyped, some are genuinely useful, and a few are actively dangerous to creative quality. I’ve been testing a range of these tools and talking to Australian post-production houses about what’s actually making it into their workflows. Here’s the honest rundown.

Where AI Is Genuinely Useful

Noise reduction and upscaling. This is probably the most mature application of AI in post. Tools like Topaz Video AI and DaVinci Resolve’s neural engine features do a genuinely impressive job of cleaning up footage shot in low light or upscaling from lower resolutions. For indie filmmakers working with limited equipment, this is a real benefit. I’ve seen AI noise reduction save footage that would have been unusable five years ago.

Rotoscoping and masking. Traditional rotoscoping is painstaking, expensive work. AI-powered roto tools in After Effects and Nuke have dramatically reduced the time needed for complex masking tasks. This is one area where the productivity gains are undeniable. What used to take a VFX artist hours can now be done in minutes with manual cleanup.

Audio cleanup. AI-powered audio tools for removing background noise, wind, and room tone issues are genuinely impressive. Adobe’s Podcast tools, iZotope RX, and similar products can rescue dialogue recordings that would previously have required ADR. For low-budget productions that can’t afford multiple days in a sound studio, this is significant.

Transcription and subtitling. Automated transcription has gotten very good. Not perfect, but good enough that it’s transformed subtitle creation workflows. For Australian filmmakers preparing festival submissions with subtitle requirements, AI transcription saves days of manual work.

Where AI Is Questionable

Colour grading. Several AI colour grading tools promise to match the look of famous films or automatically grade footage. In my experience, the results are mediocre at best. Colour grading is deeply creative work, and the AI tools I’ve tested produce generic, flat results that any experienced colourist would reject. They might be useful for social media content, but not for serious filmmaking.

Script analysis and editing suggestions. Some AI tools claim to analyse scripts and suggest structural improvements. I’d be very cautious here. Screenwriting is art, and algorithmic analysis of story structure tends to push everything toward formulaic Hollywood templates. If you want notes on your script, talk to a human who understands storytelling.

Deepfake-style face replacement. This exists, it’s getting better, and it’s a legal and ethical minefield. The technology is advancing faster than the regulations, and using it raises serious consent and likeness issues. Australian filmmakers should be extremely careful about incorporating any face-synthesis technology into their work.

The Cost Reality

Most AI post-production tools operate on subscription models. A typical indie filmmaker might spend $50-$200 per month on AI tools during post-production. That’s not nothing, but compared to hiring additional crew or booking more time at a post house, it can be cost-effective.

The bigger cost consideration is time. Learning to use these tools effectively takes time, and the output still requires human oversight and correction. The “AI will do it all” pitch from tool vendors is misleading. In practice, AI handles maybe 70-80% of certain tasks, and a skilled human does the rest.

Some larger Australian post-production facilities have started working with a Sydney-based firm to evaluate which tools are worth integrating and which are just noise. That kind of structured assessment can prevent a lot of wasted time and money chasing tools that don’t deliver on their promises.

What Australian Post Houses Are Using

I spoke to three mid-size post-production houses in Melbourne and Sydney about their current AI tool adoption. All three are using AI noise reduction and audio cleanup. Two are using AI-assisted roto and masking. None are using AI colour grading for client work.

The consensus was that AI tools are useful for specific, well-defined tasks but aren’t replacing creative decision-making. One colourist put it well: “The AI does the boring parts faster, which gives me more time for the creative parts.” That feels like the right framing for where we are in 2026.

Looking Ahead

The AI tools that will matter most for Australian filmmakers over the next couple of years are the ones that reduce the cost and time of technically complex but creatively straightforward tasks. Better noise reduction, faster roto, improved audio cleanup. The tools that try to automate creative decisions will continue to disappoint because filmmaking, at its core, is a human endeavour.

Use the tools that work. Ignore the hype. And always trust your eyes and ears over an algorithm’s suggestions.