S.M.A.C.C. NEWS DESK

In this edition:

• YouTube is reducing the reach of copied and recycled Shorts
• Sony has demanded the removal of more than 260,000 AI-generated tracks
• More than half of new music uploaded to Deezer is now AI-generated
• Brazil is discovering how difficult it is to define “AI-generated” political content
• A TikTok case in Florida raises serious questions about private information becoming public content

The creator economy never stands still.

This time, several apparently unrelated stories point towards the same increasingly important question:

As technology makes content easier to create, copy, manipulate and distribute, how do we decide what is original, what is legitimate — and who should be responsible?

Here is what caught the attention of the S.M.A.C.C. News Desk.

YouTube wants original creators to get the audience

YouTube is changing the recommendation system behind Shorts to give greater priority to original content and reduce the distribution of videos simply re-uploaded from other creators without meaningful additional contribution.

That distinction is important.

YouTube isn’t necessarily saying that the copied Short must be removed.

Instead, it can simply stop recommending it.

Channels primarily aggregating or reposting other people’s videos without significant changes are therefore likely to see less distribution through the Shorts feed.

Creators using somebody else’s clips aren’t automatically excluded. YouTube says creators should add genuine value through their own commentary, editing, storytelling or distinctive perspective.

This isn’t quite the same thing as copyright.

You might even have permission to use somebody else’s content and still discover that YouTube considers your version insufficiently original to deserve significant recommendation.

For creators, that produces an important new reality:

A platform doesn’t have to delete your content to make it virtually invisible.

And for people spending considerable time and money producing original material, perhaps there is another side to the argument.

Why should somebody who simply downloads that work and uploads it again receive the same opportunity to reach an audience as the person who created it?

For original creators, YouTube’s answer increasingly appears to be: they shouldn’t.

More than half of new music uploaded to Deezer is now AI-generated

Think about that figure for a moment.

More than half.

French streaming platform Deezer says more than half of the new music being uploaded to its service is now entirely AI-generated.

It says around 90,000 AI-generated tracks are arriving every day.

Meanwhile, Sony Music says that by the end of September it had requested the removal of more than 260,000 AI-generated tracks that it says impersonated or otherwise infringed its artists.

That number has almost doubled in approximately six months.

Artists affected reportedly include current stars as well as historic performers, with AI increasingly capable of reproducing recognisable voices and musical styles.

But imitation isn’t the only problem.

AI makes it possible to manufacture enormous quantities of music extremely cheaply and then potentially use automated listening to generate fraudulent streams and royalties.

Deezer has previously reported extraordinarily high levels of fraudulent streaming associated with fully AI-generated tracks.

The implications extend far beyond music.

Photographers, writers, designers, filmmakers, YouTubers and social-media creators should all be asking the same question:

What happens when the amount of content available becomes effectively infinite?

Until recently, creators competed against other creators.

AI changes that equation.

A human creator may eventually be competing for attention against millions — potentially billions — of pieces of machine-generated content that cost almost nothing to produce.

The scarce commodity will no longer be content.

It will be human attention.

When does using AI to edit something make it a deepfake?

Brazil is providing a fascinating demonstration of why apparently simple rules about AI can become extremely complicated in practice.

During its election period, Brazil’s Superior Electoral Court has imposed stringent restrictions concerning AI-generated or AI-altered depictions of candidates and public figures.

That sounds relatively straightforward when we’re talking about somebody generating a convincing fake video showing a politician saying something they never said.

But modern video editing isn’t that simple.

Applications including Adobe Premiere and CapCut increasingly contain AI tools as part of completely ordinary editing workflows.

Background removal can involve AI.

Audio enhancement can involve AI.

Image manipulation can involve AI.

Subtitling can involve AI.

And files can contain technical information indicating that AI tools have been involved somewhere in their creation.

Meta says its systems can automatically detect some of those indicators.

The result has been posts involving Brazilian government ministers, political activists and commentators being removed or restricted as platforms attempt to comply with electoral rules.

This creates a very important distinction.

Using AI in producing a video isn’t necessarily the same thing as using AI to deceive somebody.

If I clean the sound on a genuine interview using AI, have I produced AI-generated political content?

If I remove the background behind a politician, have I?

What if I alter their words?

What if I generate their voice?

Somewhere along that spectrum ordinary editing becomes synthetic representation.

Finding precisely where that line sits may become one of the most important problems facing regulators — and creators.

When private information becomes “content”

A very different case in Florida raises another issue creators increasingly need to understand.

TikTok creator Mark Stephen Asea was arrested on 30 September and faces a felony computer-related charge.

Police allege that private information belonging to women — including private conversations, device information and banking information — subsequently appeared in TikTok content.

One complainant told police that some of the information had only been shared within a private TikTok group chat and that she had not authorised its publication.

Asea has been accused, not convicted, and those allegations must therefore be treated as allegations unless and until established through the legal process.

But the case raises a much broader issue that applies to creators everywhere.

Creators increasingly receive:

screenshots, leaked messages, recordings, private conversations, photographs, documents and allegations about other people.

And sometimes the temptation is obvious:

That’s a great story.

But receiving information doesn’t automatically give somebody an unrestricted right to publish it.

Privacy, data protection, confidentiality, copyright, harassment and defamation can all potentially become relevant.

“Someone sent it to me” isn’t necessarily a legal or ethical defence.

As independent creators increasingly perform roles once associated primarily with journalists and publishers, they increasingly encounter some of the same responsibilities.

The bigger picture

These stories look very different.

YouTube is deciding what deserves an audience.

AI is producing music faster than humans could ever listen to it.

Governments are trying to distinguish ordinary AI-assisted editing from political deception.

And individual creators increasingly have access to information and audiences once available mainly to established media organisations.

But there is a common thread.

The power to create and publish has become extraordinarily easy to obtain.

The difficult part is deciding what responsibilities come with that power.

For creators, originality, authenticity, consent, privacy and responsibility are rapidly becoming just as important as cameras, editing software, followers and views.

The technology will continue changing.

The harder question is whether the rules — and those of us using it — can keep up.

S.M.A.C.C. — Social Media and Content Creators Community

Following the developments affecting people who create, publish and build businesses through digital media — from platform changes and monetisation to AI, copyright, regulation and creator rights.

@YouTube · @Google · @Sony Music Entertainment · @Deezer · @Meta · @TikTok · @Adobe