Guidelines for Labeling AI-Generated Content

At RSS.com, we encourage creators to be fully transparent with their audience. Labeling your episodes with the AI tag is more than a formality; it preserves listener trust, improves platform safety, and aligns with evolving industry norms and advertiser expectations.

When listeners know whether content is AI-generated, even partially, they can make informed judgments, which builds your long-term credibility. Furthermore, disclosing AI usage helps advertisers evaluate and trust your content, ensuring a healthy ecosystem for your brand.

When should you use the AI tag?

The tag should be used when the core substance of your content, whether that is the voice, the music, or the synthesis of the episode, is significantly produced by artificial intelligence.

You can mark an episode with the AI tag using the setting next to the Explicit tag:

Use the following examples to help decide if your episode requires the tag:

Case 1: Creative Assistance (Tag Not Required)

If you use AI purely as a "digital assistant" for administrative, brainstorming, or aesthetic tasks, you do not need to use the AI tag.

Examples: Using ChatGPT to brainstorm episode titles, using an LLM to fix the grammar in a script you wrote yourself, or using Midjourney to generate cover art.

Verdict: No Tag Needed. The "soul" and primary narrative of the episode are still human-led.

Case 2: Voice Cloning of Self (Optional, highly recommended)

If you write the script entirely yourself but use an AI tool to clone your own voice to read it.

Examples: You wrote a 20-minute essay but didn't have time to record it, so you used a high-quality clone of your own voice (via tools like ElevenLabs) to "perform" the reading.

Verdict: Optional (User Discretion). This sits in a gray area because the thoughts are yours, but the voice is synthesized. Because AI voices can closely mimic human styles, we highly recommend tagging this to maintain ultimate transparency and protect listener trust.

Case 3: Fully Synthetic or AI-Scripted Content (Tag Required)

If the primary content, either the written script, the audio narration, or both, is generated by AI with minimal human intervention.

Examples: Using ChatGPT or another LLM to write the actual script of your podcast, using NotebookLM to turn a PDF into a deep-dive conversation, or having a generic AI narrator read your content.

Verdict: Tag Required. Without disclosure, listeners might be misled about what is human versus AI. Listeners and advertisers must know the insights or voices were generated by a machine.

Case 4: Audio Post-Production (Tag Not Required)

If you recorded a standard human podcast but used AI strictly to "clean up" the audio file.

Examples: Using "Enhance Speech" tools (like Auphonic) to remove background noise, or using AI-based tools to cut out filler words like "umms" and "ahhs."

Verdict: No Tag Needed. These are considered standard editing tools, acting similarly to a traditional volume knob or an equalizer.