When I started writing my book, Survival of the Artist, artificial intelligence was not an everyday tool for most people.
Now it is everywhere.
Artificial intelligence (AI) is technology that can analyze information and generate things such as text, images, audio and ideas based on instructions from a user.
ChatGPT and other generative AI tools, which create new content from prompts, can solve problems quickly and help with things you may not know how to do yet.
I use AI too.
Still, the biggest lesson has not changed since I first wrote this article: AI can help you start. It should not replace the part that makes the work yours.
Use AI as an assistant, not the artist
When I first played around with ChatGPT, I asked it for an outline of a book about the importance of social media for artists.
It gave me some terrific information. Some of it was already in my book, while other ideas gave me new directions to think about.
What it could not provide was my personal touch, tone and life experience.
That distinction matters even more now.
AI tools have become much better at producing believable writing, images, voices and music. At the same time, streaming services, distributors and the rest of the music business have become much more careful about AI-generated content, impersonation and mass-produced music.
So how can an artist use AI without handing over the creative process?
1. Build a marketing plan
Creating a marketing plan can be one of the hardest parts of releasing music.
You know the songs. You may not know what to post six weeks before release, what information you need from your team or how all the pieces should fit together.
AI can help you create a starting outline.
For example, give it the release date, genre, audience, available budget and the platforms you actually use. Then ask it to organize possible marketing tasks into a timeline.
Afterward, edit the plan based on what you know about the artist.
AI does not know your audience better than you do.
If you are still figuring that audience out, our guide to defining your target audience is a better place to start.
2. Get past the blank social media post
Creating social copy can be draining, especially when you have already written 12 versions of “our new song is out Friday.”
AI can help brainstorm angles, captions or questions.
Then make them sound like you.
If the finished post sounds like a marketing department generated it for an artist nobody has ever met, start over.
The point is not to let AI speak for you. Instead, use it when your brain needs help finding another way into the idea.
Our guide to creating social content people actually care about can help you decide what is worth posting in the first place.
3. Start bigger writing projects
Press releases, biographies, project proposals and other professional documents can feel intimidating when the page is completely blank.
AI can give you a structure to work from.
However, give it real information rather than asking it to invent the story for you.
Provide verified facts, dates, names, accomplishments and quotes. Then check the finished draft against your source material.
AI can confidently write something that sounds completely reasonable and is completely wrong.
That is sometimes called an AI hallucination, which happens when an AI system generates false or unsupported information as though it were factual.
So verify everything before publishing it.
If the project is a press release, our guide to the basics of writing a press release gives you the structure AI should be helping with, not replacing.
4. Use AI for visual ideas
AI image tools can be useful when you need to explore a visual direction quickly.
Maybe you want to test concepts for a photo shoot, build a mood board, experiment with layouts or create a rough mockup before spending money on the final design.
That can save time.
Still, be careful about using generated images that imitate another artist, photographer or recognizable person’s identity. Also make sure anything used commercially fits the rights, platform rules and agreements surrounding your project.
For actual release artwork, you still need something that represents the artist and meets distribution requirements. Our guide to music cover art covers those basics.
5. Let AI handle some of the boring stuff
This may be my favorite use.
Not every AI task needs to be creative.
You can use it to help:
- organize notes
- turn a list into a schedule
- summarize material you already have
- brainstorm interview questions
- clean up repetitive information
- create checklists
- compare versions of a document
- explain a concept you are trying to learn
These are the jobs that can eat an hour without adding anything particularly creative to your day.
Use the machine for machine work.
6. Be much more careful with AI music and voices
This is where things have changed the most since the original version of this article.
AI can now generate complete songs, instrumentals and highly convincing versions of real people’s voices.
Technically impressive does not mean commercially safe.
An AI voice clone is a synthetic voice designed to imitate a real person’s recognizable voice. Using someone’s voice or identity without authorization can create serious rights and platform problems.
Spotify, for example, now expressly prohibits unauthorized vocal impersonation and has added AI disclosures and other tools intended to distinguish AI use from authentic artist activity.
Deezer has gone even further by identifying fully AI-generated music, labeling it and excluding it from algorithmic recommendations.
Syntax Creative does not accept AI-generated music for distribution.
That does not mean every use of AI anywhere in the creative process is the same thing.
An artist might use technology to explore chord changes, clean up a workflow, experiment with an idea or solve a technical problem. The important distinction is whether the technology is helping the artist make the music or becoming the artist.
My original advice included this line, and it still works:
Maybe you need help hitting a harmony. Fine.
Just do not make that high harmony sung by Michael Jackson.
AI-generated music is no longer a hypothetical problem
When this article was first published in 2024, fully AI-generated artists and songs still felt somewhat experimental.
That changed quickly.
By 2026, Deezer reported that AI-generated music had exceeded half of its new daily uploads at its peak.
Spotify has also introduced AI-related credits, stronger impersonation policies and labels for artist identities that appear to be AI-generated.
In other words, the music business is no longer asking whether AI-generated music might arrive.
It is figuring out what to do with it now that it is here.
Do not outsource your judgment
AI can produce an answer in seconds.
That does not mean it produced the right answer.
Check facts. Read the final copy. Listen critically. Make sure you understand what you are putting your name on.
Most importantly, do not let convenience remove the part of the process where you make decisions.
Your taste is part of the work.
Let AI enhance what you do, not replace it
There are a lot of great AI tools available now, and there will probably be another dozen by the time you read this.
You do not need all of them.
Use AI for mundane tasks, brainstorming, learning and getting past the blank page. Then tweak it, adjust it and breathe some life back into it.
You should still understand how to do the work yourself.
Eventually, you may hire a writer, designer, marketer, producer or another professional who can do the job better than the machine.
Until then, AI can be a useful assistant.
Just do not confuse the assistant with the artist.
This article is adapted from Justin Sarachik’s book, Survival of the Artist: Tips & Strategies to Help Indies Succeed.
This article provides general information and is not legal advice. Artificial-intelligence technology and the laws and platform policies surrounding generated content, copyright, voice, likeness and disclosure continue to evolve.
