Tidal AI Policy: Demonetization of AI-Generated Music

Tidal has implemented a new AI policy to combat the proliferation of AI-generated content and ensure that royalties are prioritized for human creators. The core of the policy is the demonetization of fully AI-generated music, removing the financial incentive for "slop" content to flood the platform.

Demonetization and Monetization Standards

Tidal will no longer allow music that is 100% AI-generated to be monetized. These releases will not receive royalties, nor will they be eligible for direct-to-fan sales. This move aligns with the legal perspective that works created solely by AI may not be copyrightable and should therefore be treated as public domain content.

Content Integrity and Artist Protection

Tidal is introducing strict standards for content integrity to prevent fraud and impersonation. The platform will not tolerate AI-generated music that:

• Impersonates a specific artist or group.
• Exploits an individual's name or likeness.
• Deceives listeners or facilitates fraudulent activity.

To enforce these rules, Tidal is deploying automatic tools to identify and remove impersonating releases immediately and on an ongoing basis.

Transparency and Labeling

To inform listeners, Tidal will identify and tag music detected as wholly AI-generated with an "AI" badge within the app. This transparency measure is intended to help users distinguish between human-authored works and generative AI outputs.

Implementation and Technical Challenges

The policy changes took effect on July 15. While Tidal aims to expand these rules to music that is "substantially AI-generated," they have stated they will only do so once AI detection technology becomes sufficiently reliable.

Community discussion highlights several critical challenges regarding this implementation:

The Detection Gap

Technical experts argue that detecting AI-generated audio is a "cat and mouse game." As AI models improve, the "tells" used by detection software decrease, leading to concerns about false positives and false negatives. Some users suggest that without self-reporting from uploaders, automated detection at the audio layer is an LSTM-hard problem.

Definition Ambiguity

There is significant debate over what constitutes "substantially AI-generated." Users have questioned how the policy will apply to hybrid workflows, such as:

• Lyrics written by LLMs but music composed by humans.
• Human vocals performed over an AI-generated backing track.
• AI vocals layered over human-instrumented tracks.

User Experience and Market Demand

While some users appreciate the AI-generated music for specific moods (e.g., vaporwave or outrun), others express a strong desire for a full opt-out mechanism to remove AI content from their feeds entirely. Some subscribers have reported that their feeds are already being flooded with AI-generated releases that spoof the names of artists they follow, making the policy a necessary but overdue step.