Every language has a song worth hearing.
The Songumentary Global AI Music Awards bring Melbourne's language communities and creators from around the world together to make songs, poems and short documentaries with AI. It is one public day of making, showing and voting, with an online round open to everyone.
What it is
Most AI music awards reward the loudest English-language releases. This one starts from the other side: the many languages Melbourne already speaks, and the stories their communities carry.
Creators make a song, a poem or a songumentary, a true story told in song. Some of the best work will carry a story across languages, with AI helping to adapt it from one culture to another.
It runs as an online round open worldwide, then one public day in Melbourne: a morning creation challenge, an afternoon showcase where the audience votes, and an evening awards night. Every entry plays with live on-screen transliteration and translation, so a mixed audience can follow all of it.
Vision and values
Our vision. Every community can tell its own story, in its own language, and have it heard by people who do not speak it. AI is a bridge between cultures. People stay at the centre: the story, the judgement and the credit belong to them.
Community first
Melbourne's language communities shape the awards: the categories, the juries and the audience.
Many languages, one stage
No language is a side category. Each has its own award and its own jury.
Honest about AI
Everyone says how AI was used, and nobody is penalised for saying it.
First Nations first
First Nations stories lead their own category and award, on First Nations terms.
People decide
People judge, people check translations, and people are credited for the work.
Craft and story
Work is judged on what it says and how well it is made, not on how much or how little AI touched it.
Open to everyone
Solo creators, teams, elders, young people. Auslan, interpreters and captions.
Resilience
AI creators meet doubt and insults. We answer with good work and open rules, not with insults.
What the first AI Music Awards in Los Angeles taught us: read the lessons.
Award categories (draft)
Cultural bridging
Best cross-language song adaptation. Best poetry adaptation. Best community collaboration.
Songumentary
Best songumentary. Best social voice: an issue or a lived experience told in song.
First Nations
Best First Nations story and a First Nations artist award, to be judged by a First Nations panel.
Craft
Lyrics, voice and performance, music video, production.
Open
Best human and AI together. Comedic or light song. Youth award. Elder or heritage award.
By language
One "Best Song in…" award for each language.
Special
Audience Choice. Hackathon Team Prize. Spirit of Songumentary.
Categories are a draft. The advisory team will set them.
Languages
About ten languages, chosen with Melbourne's language communities and broadcasters. We begin with:
EnglishHindiGreekItalian
The full list will be announced with the advisory team.
The day in Melbourne
Morning
Briefs are revealed on stage. Teams of one to four pick a story to tell: fiction, essay or social issue, all welcome.
Afternoon
Teams finish a short song, poem or songumentary and present it live. The audience votes.
Evening
The awards night, with finalists from the online round, livestreamed worldwide.
How we use AI, and our ethics
This is our starting position. The advisory team, including First Nations cultural advisers and a rights adviser, will review it before entries open.
- Disclosure, not detection. Every entry says how AI was used: which tools, what a person did, what AI did. We group entries by that declaration so like is judged with like.
- People in charge. AI does not judge or choose winners. Juries are people, drawn from each language community.
- No impersonation. No cloning or imitating the voice or likeness of a living person without their written consent.
- First Nations material. We follow Indigenous Cultural and Intellectual Property principles. No AI generation of First Nations languages, voices, or sacred or restricted material without the consent of the relevant community.
- Accuracy across languages. Speakers of each language check the translations and transliterations shown at the event.
- Adaptation and credit. An adapted song or poem credits the original creator and the adapter. Entrants adapt their own work, public-domain work, or work they have written permission to adapt.
- Tools and licences. Entrants are responsible for having the right to use their tools' output in a public work. We do not endorse any tool.
- Your entry stays yours. We do not train AI on entries.
- Fair judging. Declared AI use is not a penalty. Rules, jurors and conflicts of interest are published before entries open.
- Honest conversation. We welcome serious critique of AI in music, including from people who object to it. We do not answer abuse with abuse.
See also AkunaMedia's approach to AI and craft.
Join the team
We are building the team in three stages. Tell us which one fits you.
1. Advisory team
Broadcasters, language-community leaders, First Nations cultural advisers, AI music practitioners, rights experts. A light commitment. Your judgement shapes the rules, categories and juries.
2. Working team
Language leads, hackathon and technical leads, voting and website, community outreach, communications, accessibility and operations.
3. Leadership team
Executive producer, artistic director, First Nations director or co-chair, finance and compliance. Named once the event is funded.
Get in touch
Whether you want to advise, help run it, enter, come along, sponsor or cover it, tell us below. You can also email festival@songumentary.com.
Timeline
- November 2026
Funding application to Now or Never 2027. Advisory team begins.
- February 2027
Funding outcome.
- March to April 2027
Working team and language leads confirmed. Rules published. Online round opens.
- June to July 2027
Venue confirmed. Online round closes and finalists are named.
- Late August 2027
Hackathon, showcase and awards in Melbourne.
Dates depend on funding and may change.