AI Model Usage Policy

This AI Model Usage Policy governs your use of AI models provided through TaskAGI. This policy supplements our Terms and Conditions and Acceptable Use Policy. By using any AI models on TaskAGI, you agree to these terms.

1.1. Third-Party AI Model Providers: TaskAGI integrates 30+ AI models from third-party providers, including but not limited to:

• Text Generation: OpenAI (GPT-4, GPT-3.5), Anthropic (Claude), Google (Gemini), Meta (Llama)
• Image Generation: DALL-E, Midjourney, Stable Diffusion, Replicate models
• Audio/Voice: ElevenLabs, HyperVoice (TaskAGI proprietary), OpenAI TTS, Google Text-to-Speech
• Video Generation: RunwayML, Replicate video models
• Computer Vision: Google Vision, OpenAI Vision, specialized detection models

1.2. Model Availability: AI models may be added, removed, or modified at any time without notice. Some models may have usage quotas or rate limits.

1.3. Third-Party Terms: When using third-party AI models, you are also bound by the provider's terms of service. It is YOUR responsibility to comply with:

• OpenAI Usage Policies: https://openai.com/policies/usage-policies
• Anthropic Acceptable Use Policy: https://www.anthropic.com/legal/aup
• Google Gemini Terms: https://ai.google.dev/gemini-api/terms
• Other provider-specific terms as applicable

⚠️ READ THIS SECTION CAREFULLY - YOU ARE SOLELY RESPONSIBLE FOR ALL AI-GENERATED CONTENT

2.1. No Accuracy Guarantees:

• AI models can and DO produce inaccurate, misleading, or completely false information ("hallucinations")
• AI outputs may contain factual errors, outdated information, or nonsensical content
• TaskAGI makes NO warranties about the accuracy, reliability, or fitness of AI-generated content
• YOU MUST verify all AI outputs before use in any critical application

2.2. No Intellectual Property Guarantees:

• AI models may generate content that infringes third-party copyrights, trademarks, or other IP rights
• TaskAGI does NOT warrant that AI outputs are original or non-infringing
• YOU are responsible for ensuring AI outputs do not violate third-party IP rights
• YOU may be held liable for copyright infringement if you use AI-generated content that violates IP laws

2.3. No Professional Advice:

• AI outputs do NOT constitute professional advice (legal, medical, financial, etc.)
• AI models are NOT licensed professionals and cannot replace qualified experts
• Reliance on AI-generated professional advice is at YOUR OWN RISK
• Always consult qualified professionals for critical decisions

2.4. User Review Obligation:

• YOU ARE SOLELY RESPONSIBLE for reviewing, editing, and approving all AI-generated content before use
• YOU must verify factual claims, check for bias, and ensure content meets your requirements
• TaskAGI is NOT liable for any consequences of using AI outputs without proper review

2.5. Bias and Harmful Content Risks:

• AI models may produce biased, discriminatory, offensive, or harmful content
• Models reflect biases present in their training data
• YOU must review outputs for bias, fairness, and appropriateness
• TaskAGI is not responsible for harmful content generated by AI models

You may NOT use TaskAGI AI models for:

3.1. Illegal Activities:

• Generating content that violates laws (fraud, defamation, harassment, threats, etc.)
• Creating deepfakes or impersonations without consent
• Circumventing security measures or access controls
• Any activity prohibited by applicable law

3.2. Harmful Content Generation:

• Child sexual abuse material (CSAM) or child exploitation content
• Non-consensual intimate images or revenge pornography
• Content promoting violence, terrorism, or extremism
• Self-harm or suicide promotion
• Hate speech or incitement to discrimination

3.3. Deceptive Practices:

• Generating spam, phishing, or scam content
• Creating fake reviews, testimonials, or social media engagement
• Impersonating individuals or organizations without disclosure
• Manipulating elections or political processes
• Academic dishonesty or plagiarism

3.4. High-Risk Decision-Making (Without Human Oversight):

• Medical diagnosis or treatment recommendations
• Legal case outcomes or sentencing recommendations
• Employment decisions (hiring, firing, promotions)
• Credit or insurance eligibility determinations
• Law enforcement decisions

Note: AI may assist in these areas WITH qualified human oversight and final decision-making.

3.5. Unauthorized Scraping/Training:

• Using AI to scrape websites that prohibit AI scraping in their ToS or robots.txt
• Training custom AI models on data you don't have rights to use
• Reverse-engineering or extracting training data from AI models

4.1. Credit-Based System: AI model usage consumes AI credits from your account balance. Each model has a different cost per operation (see integration details).

4.2. Non-Refundable Credits: AI credits are non-refundable once used, even if you are dissatisfied with AI outputs. See our Refund Policy.

4.3. Rate Limits: We may impose rate limits or usage quotas to prevent abuse and ensure fair access. Excessive usage may result in temporary suspension.

4.4. Provider Cost Changes: If third-party AI providers increase their pricing, we reserve the right to adjust AI credit costs with 30 days' notice.

5.1. Input Data: When you use AI models, your input prompts and data are sent to third-party AI providers for processing.

5.2. Provider Data Policies:

• OpenAI: API usage is NOT used to train models (as of March 2023). See OpenAI Data Policy
• Anthropic: Does not train on user data. See Anthropic Data Policy
• Google: May use data to improve services unless you opt out. See Google Privacy Policy
• Other providers: Check provider-specific data policies

5.3. Sensitive Data Warning: DO NOT send sensitive personal data, confidential information, or trade secrets to AI models unless you understand and accept the provider's data handling practices.

5.4. GDPR Considerations: If you process personal data through AI models, you are the Data Controller and responsible for GDPR compliance. See our Data Processing Agreement.

6.1. Provider Safety Filters: Third-party AI providers implement safety filters to block harmful prompts and outputs. TaskAGI does NOT override these filters.

6.2. Blocked Requests: If your prompt violates a provider's safety policies, the request will be blocked, and you may still be charged credits.

6.3. Monitoring for Abuse: We reserve the right to monitor AI usage for violations of this policy and our Acceptable Use Policy.

6.4. Account Suspension: Repeated violations may result in account suspension or termination without refund.

7.1. No Liability for AI Outputs: TaskAGI is NOT liable for:

• Inaccurate, false, or misleading AI-generated content
• Intellectual property infringement by AI outputs
• Harmful, biased, or offensive AI-generated content
• Losses resulting from reliance on AI outputs
• Damages caused by using AI outputs without proper review

7.2. User Indemnification: YOU agree to indemnify and hold TaskAGI harmless from:

• Claims arising from your use of AI-generated content
• IP infringement claims related to AI outputs you published
• Harm caused by AI content you created through TaskAGI
• Violations of third-party AI provider terms
• Any prohibited uses of AI models

7.3. Maximum Liability: To the extent permitted by law, TaskAGI's total liability for AI-related claims is limited to the amount you paid for AI credits in the 3 months prior to the claim.

8.1. Model Updates: Third-party providers may update AI models at any time, which may change output quality, behavior, or capabilities.

8.2. Model Removal: We may remove AI models from TaskAGI if:

• The provider discontinues the model
• Pricing becomes commercially unviable
• The model violates our policies or legal requirements

8.3. No Migration Guarantees: If a model is removed, we are NOT obligated to migrate your workflows to alternative models. You are responsible for updating your workflows.

8.4. 30-Day Notice: We will provide 30 days' notice before removing popular AI models, where possible.

9.1. Who Owns AI Outputs?

• You retain ownership of your input prompts
• Ownership of AI outputs depends on the provider's terms:
  • OpenAI: You own outputs (subject to usage policy compliance)
  • Anthropic: You own outputs
  • Others: Check provider-specific terms
• TaskAGI does NOT claim ownership of AI outputs you generate

9.2. IP Infringement Risk: Even if you "own" AI outputs per provider terms, this does NOT mean outputs are guaranteed to be non-infringing. You are still responsible for ensuring outputs don't violate third-party IP rights.

9.3. Attribution: Some AI providers require attribution when publishing AI-generated content. Check provider-specific requirements.

10.1. AI Disclosure: In certain contexts, you may be legally required to disclose that content was AI-generated:

• Deepfakes or synthetic media (required in many jurisdictions)
• AI-generated news articles (transparency requirements)
• Political advertising (election laws)
• Academic work (institutional policies)

10.2. User Responsibility: YOU are responsible for complying with applicable disclosure laws. TaskAGI does not automatically add AI disclosure labels to outputs.

11.1. Restricted Countries: AI models may not be available in countries subject to EU, UK, or US sanctions.

11.2. Export Compliance: You may not use AI models to develop weapons, nuclear technology, or other restricted applications under export control laws.

We may update this AI Model Usage Policy at any time. Material changes will be notified via email or platform notice. Continued use after changes constitutes acceptance of the updated policy.

AI Policy Questions: support@taskagi.net
Report AI Abuse: abuse@taskagi.net
Legal Inquiries: legal@taskagi.net

Related Documents:

• Terms and Conditions
• Acceptable Use Policy
• Privacy Policy
• Data Processing Agreement