Kunda Kids AI Usage Policy

1. Introduction & Core Philosophy

This policy governs the responsible, secure, and ethical use of Artificial Intelligence (AI) tools by Kunda Kids staff. Given our focus on children's media, edtech, and publishing, our primary obligations are to maintain the highest standards of Child Safety, Cultural Integrity, and Intellectual Property (IP) protection.

Kunda Kids AI Philosophy: Partner, Not Magician

AI tools should be treated as creative and process partners, not as end-to-end solutions.

• Originality Mandate: AI must be used to assist in the creation of original ideas, not to plagiarise or generate complete, unedited, final works. Every piece of final content must retain the distinct voice and vision of Kunda Kids.

• Human-Led Creation: The core creative and strategic input must always originate from the Kunda Kids staff member. The human user must be the architect of the content, with AI serving as the builder, editor, or brainstorming assistant.

• Full Ownership: AI should only be used in a manner that ensures Kunda Kids retains full, exclusive, and unencumbered ownership of the resulting content and ideas (see Section 3).

2. Data Security & Confidentiality Protocols

The following data must NEVER be input into any third-party or Public Generative AI Tool (e.g., public LLMs, image generators) as the data may be used to train the model, compromising our IP and confidentiality:

• Confidential IP / Trade Secrets: Unpublished manuscripts, story outlines, character designs, music/lyric drafts, proprietary algorithms, financial models.

• Personal & Sensitive Data: Customer/student data, staff PII, vendor contracts with identifiable information, or confidential communications.

• Proprietary Code: Kunda Kids edtech platform source code, database schemas, API keys, or security credentials.

Crucial Note: If an AI tool's terms of service are ambiguous regarding data retention or the use of user input for model training, do not use it with Kunda Kids data.

3. Intellectual Property (IP) and Ownership

3.1. Ownership of Output

• Any output generated using AI tools during Kunda Kids' employment is considered Work for Hire. All resulting IP (copyright, trademarks, etc.) belongs exclusively to Kunda Kids.

• Record Keeping: Staff must retain records of the specific tool used and the prompt for any key piece of AI-generated content (e.g., for subsequent IP audits).

3.2. Copyright Risk and Infringement

• Prohibited Prompts: Staff must strictly avoid prompts that request the AI to mimic, imitate, or generate content “in the style of” a specific, named contemporary creator or copyrighted IP.

• Verification: All AI-generated text used for final published works must be thoroughly checked for factual accuracy and plagiarism.

4. Content Quality and Ethical Guidelines

4.1. Accuracy and Fact-Checking

• AI is prone to "hallucinations." Any factual statement, data point, or citation generated by AI and intended for public release or edtech materials must be independently verified using reliable human-curated sources.

• Verification is mandatory before any AI-assisted content is signed off.

4.2. Cultural Sensitivity and Bias Mitigation

• Staff must critically evaluate all AI outputs for inherent biases, stereotypes, or culturally inappropriate representations, especially concerning race, ethnicity, gender, and African heritage.

• Prompt Engineering: Use specific, inclusive, and culturally-aware prompts to ensure the AI output aligns with the authentic and positive narratives of Kunda Kids.

5. Approved and Prohibited AI Tools

Responsibility for tool approval rests with the relevant Team Leads, adhering strictly to the Data Security (Section 2)and IP Ownership (Section 3) requirements.

Process Note: Any team member wishing to use a new AI tool must submit a request to their Team Lead for review against this policy before use.

6. Training, Review, and Compliance

6.1. Attribution and Transparency

• Internal Disclosure: When sharing AI-assisted work internally, label it with a clear disclaimer, such as "AI-Assisted Draft – Requires Human Review," and specify the tool used.

• External Disclosure: All AI-assisted content released to the public must adhere to Kunda Kids' transparency standards (e.g., explicit disclosure in a byline, credit, or colophon) or as defined by the Chief Operations Officer /Team Lead.