AI LIMITATIONS & ASSUMPTION OF RISK

Effective Date: August 22, 2026

This page applies to any engagement involving our Artificial Intelligence services — including AI and AI Agent Development, AI Automation, LLM Development, Custom AI Models, Dataset Engineering, AI Research & Theorems, and AGI Strategy, Research & Forecasting — and to any deliverable that incorporates AI-generated or AI-assisted output.

1. Nature of AI Technology

Artificial intelligence and large language model technologies are probabilistic systems. They can produce output that is inaccurate, incomplete, outdated, biased, or entirely fabricated (sometimes called “hallucination”), even when the output appears confident or well-formatted. This is a known limitation of the underlying technology, not a defect specific to our work.

2. No Guarantee of Outcomes

We do not guarantee the accuracy, reliability, completeness, or fitness for a particular purpose of any AI-generated or AI-assisted output, model, agent, or automation we build, unless a specific accuracy or performance benchmark is explicitly stated and agreed in your signed SOW. This applies with particular force to AI Research & Theorems and AGI Strategy & Forecasting engagements, which are exploratory and advisory in nature given how early-stage and fast-moving this field is.

3. Your Responsibility Before Relying on AI Output

You are responsible for independently reviewing, testing, and validating any AI-generated or AI-assisted output before relying on it, publishing it, or using it in a production, customer-facing, regulated, or safety-critical context. This is especially important where output could inform:

• Legal, medical, financial, or other professional/regulated decisions
• Safety-critical systems or decisions affecting health or safety
• Public-facing communications or claims about your business

Where these contexts apply, we recommend obtaining independent professional advice rather than relying solely on AI-generated content or automation.

4. Data You Provide for AI/ML Work

If you provide data for training, fine-tuning, or grounding an AI system, you are responsible for ensuring you have the necessary rights and consents to use and share that data for that purpose.

5. Assumption of Risk

By engaging our AI-related services, you acknowledge and accept the inherent limitations and risks of AI technology described above, and agree that AI-related deliverables are tools intended to assist — not replace — human judgment and oversight.

6. Liability

This page supplements, and does not replace, the Limitation of Liability provisions in our Master Terms of Service, which continue to apply to AI-related engagements.

7. Contact Us

Questions about the scope or limitations of an AI engagement can be sent to legal@navritu.digital.