AI Content Disclosure Policy

Effective Date: May 6, 2026  ·  Last Updated: September 22, 2026

1. Policy Overview

VocalMax ("VocalMax," "we," "us," or "our"), operated by Brilliant5 Inc., is an artificial intelligence service that generates spoken audio from text using machine learning models. This AI Content Disclosure Policy ("AI Disclosure Policy") explains:

• How VocalMax uses AI to create audio content.
• Which third-party AI providers we rely on.
• Your obligations to disclose AI-generated audio when you publish it.
• How we comply with current and emerging AI transparency laws.

This policy supplements our Terms of Service and Privacy Policy. By using VocalMax, you acknowledge that audio created through our Service is AI-generated and accept the disclosure obligations described in Section 6.

2. What Is AI-Generated Audio?

VocalMax uses artificial intelligence to convert text you provide into synthetic speech. The audio output:

• Is NOT recorded by a human voice actor reading the text.
• Is generated by machine learning models trained on large datasets of human speech.
• May use voices that are entirely synthetic, or voice clones derived from human samples (with appropriate consent).
• Is rendered through automated text-to-speech and prosody modeling, not human performance.

Audio generated by VocalMax constitutes "synthetic content," "AI-generated content," "deepfake" (in regulatory definitions covering synthetic voice), or "generative AI output" under various legal frameworks.

3. Third-Party AI Providers

VocalMax relies on the following third-party AI service providers to deliver our Service:

3.1 ElevenLabs Inc. — Text-to-Speech and Voice Cloning

• Provides the underlying neural text-to-speech models that convert your text into spoken audio.
• Provides voice cloning capabilities used in our voice cloning feature.
• Reference: elevenlabs.io and ElevenLabs Privacy Policy

3.2 How Word Allowance Works for AI-Assisted Workflows

When you generate written content with VocalMax AI Create and then convert that content to audio within 24 hours, we count the words once — at the AI Create step. The subsequent audio generation does not consume additional words from your balance, up to the word count of the original AI Create generation. Words beyond that original count (from edits or expansions) are deducted normally. This applies automatically with no setup required. This feature is designed to support efficient AI-assisted audiobook workflows and is described in detail in our Terms of Service (Section 4.6).

3.3 Direct AI Integration

We directly integrate the following AI services:

• Anthropic Claude — used for natural language processing tasks including the AI Create feature (content generation, document summarization, and other text-based AI assistance). Inputs are processed by Anthropic per their Privacy Policy and Usage Policy.
• ElevenLabs — used for voice synthesis (text-to-speech) and voice cloning. Audio inputs and text-to-be-synthesized are processed by ElevenLabs per their Privacy Policy.

3.4 Platform-Level AI Services

Our hosting platform, Base44, may use additional AI providers to deliver platform features such as app-building assistance and infrastructure orchestration. Base44 publishes the current list of AI service providers it uses in their documentation. These platform-level AI services are governed by Base44's Data Processing Agreement and the respective providers' terms.

Important Note

If you have specific requirements regarding which AI providers process your data (for example, regulatory requirements that prohibit data flowing through specific countries or providers), please review the full list of Base44's subprocessors in their Data Processing Agreement before subscribing, or contact us before subscribing.

4. How We Train and Customize AI Models

VocalMax does NOT train its own foundation AI models. We use commercially-licensed models from the providers listed in Section 3. We do, however, customize the use of these models in the following ways:

• Voice clones: When you use voice cloning, we generate a customer-specific voiceprint with your provided audio samples. These voiceprints are used only for your account.
• Prompt engineering: We craft instructions to AI models to produce VocalMax-specific results (such as audiobook chapter detection).
• Output filtering: We apply content moderation and quality checks to AI-generated outputs.

We do NOT use customer content to train general-purpose AI models. Per our agreements with our AI providers, customer content uploaded to VocalMax is not used to train provider models for the benefit of other customers.

5. Content Watermarking and Detection

Audio generated by VocalMax may include automated watermarks (audible or inaudible) embedded by the underlying AI provider's technology. ElevenLabs incorporates watermarking and other provenance signals as required by their compliance with applicable AI transparency frameworks.

We do not currently provide a public AI-content detection tool because:

• VocalMax has fewer than 1,000,000 monthly users, which is the threshold for the requirement under California SB 942 and AB 853.
• We are a deployer of AI services, not a foundation model developer.

If our user base grows above the regulatory thresholds, we will implement a public detection tool and watermarking disclosure as required by law.

6. Your Disclosure Obligations When You Publish AI-Generated Audio

When you publish, distribute, or share audio generated by VocalMax, you may be legally required to disclose that the audio is AI-generated. This obligation is YOURS, not ours. You are responsible for compliance with all applicable laws in the jurisdictions where you publish audio.

6.1 Audiobook Distribution Platforms

Major audiobook distributors require AI narration disclosure:

• Audible / ACX — requires disclosure of AI narration in metadata.
• Apple Books — accepts AI narration with required disclosure tag.
• Google Play Books — has a specific "auto-narrated" category for AI audiobooks.
• Spotify Audiobooks (Findaway Voices) — accepts AI narration with metadata flag.
• Kobo Writing Life — accepts AI narration with disclosure.

Failure to disclose AI narration when required by these platforms can result in delisting, refund obligations to listeners, or termination of your account with the platform.

6.2 EU AI Act (Article 50)

The EU Artificial Intelligence Act, which becomes enforceable for transparency obligations in August 2026, requires that synthetic content (including AI-generated audio) be clearly labeled as artificial when deployed in the European Union. Penalties may reach 1.5% of global annual turnover or €7.5 million.

6.3 California AI Transparency Act (SB 942 / AB 853)

California's AI Transparency Act, as amended by AB 853, requires Covered Providers to provide users with tools to label AI-generated content with a clear and conspicuous disclosure. The effective date is August 2, 2026. While VocalMax may not currently meet the user threshold to be a Covered Provider, you should consider voluntarily disclosing AI generation in California-distributed content.

6.4 State-Level Deepfake and Synthetic Media Laws

Multiple U.S. states (Texas, Minnesota, Washington, and others) have specific laws regarding deepfake disclosure, particularly for election-related and adult content. In Canada, synthetic-media and voice-likeness issues are addressed through evolving privacy law and personality/likeness rights rather than a single dedicated statute. Verify the law of every jurisdiction where you publish audio.

6.5 Recommended Disclosure Language

For audiobooks and other published audio, consider including the following disclosure (or platform-specific equivalent):

"This audiobook was narrated using AI-generated voice technology powered by VocalMax. The narration is synthetic and was not performed by a human voice actor."

7. Accuracy and Quality of AI Output

AI-generated audio may contain:

• Mispronunciations of unusual words, names, or technical terms.
• Incorrect emotional tone or pacing.
• Pronunciation variations between different runs of the same input text.
• Other artifacts of automated speech generation.

You are responsible for reviewing audio outputs before publishing them. We are not responsible for errors in pronunciation, intonation, or other quality issues that result from automated speech generation. Where output quality is critical (e.g., for medical, legal, or financial content), we recommend manual review and, where appropriate, human voice actor narration as an alternative.

8. AI Hallucination Disclaimer

Where VocalMax uses AI for content generation features (such as automatic chapter detection or content suggestions), the AI may produce outputs that are factually inaccurate, incomplete, or misleading. This is known as "hallucination." You are solely responsible for verifying any AI-generated content suggestions before relying on them. We do not warrant the accuracy or completeness of AI-suggested content.

9. Changes to This Policy

AI regulation is evolving rapidly. We will update this AI Disclosure Policy as laws change and as our service evolves. Material changes will be communicated to active users via email and through an in-app notification.

10. Contact

Questions about AI generation, disclosure, or this policy:

For information about our designated Privacy Officer under Quebec Law 25, see our Privacy Policy.