AI Usage and Safety Policy

1. About This Notice

This AI Safety Policy is a customer-facing notice about Chat4U's AI-related risks, safeguards, and customer responsibilities. It supplements the Terms of Service, Acceptable Use Policy, AI Services Addendum, Privacy Policy, and Data Processing Addendum.

This notice is informational. It is not a separate acceptance flow, and it does not itself authorize any use case.

---

2. Core Principles

Chat4U is designed to help customers provide AI-assisted answers using customer-authorized content and configured workflows. AI systems can produce inaccurate, incomplete, outdated, biased, unsafe, or misleading output.

Customers must test, supervise, and monitor their own use of Chat4U. Chat4U does not guarantee that outputs are correct, lawful, non-infringing, safe, or suitable for reliance. Embedded widgets display mandatory AI and privacy notices; those notices do not constitute legal clearance or a guarantee of output accuracy.

---

3. Customer Responsibilities

Customers are responsible for:

• selecting an appropriate use case;
• maintaining lawful and authorized source content;
• configuring prompts, models, guardrails, escalation paths, and fallback behavior;
• reviewing outputs before deploying sensitive or high-consequence workflows;
• providing required visitor notices and obtaining required consents;
• monitoring for hallucinations, harmful output, abuse, and prompt injection;
• disabling or revising agents, tools, or data sources when risks become unacceptable.

---

4. Retrieval, Source Content, and Knowledge Gaps

Chat4U may use retrieval-augmented generation and related indexing workflows to ground responses in customer-provided or customer-authorized content. Those mechanisms reduce some risk, but they do not guarantee completeness or correctness.

Relevant information may be missing, stale, misclassified, or not retrieved. Customers should treat answer quality, source availability, and knowledge-gap signals as operational indicators rather than guarantees.

---

5. Pro Agent and Tool Safety

Pro Agent features can read page state and request browser actions. These features require careful use because they may affect page navigation, form content, user choices, or downstream website behavior.

Customers must not configure Pro Agent to:

• collect passwords, one-time codes, payment card data, or other high-sensitivity secrets;
• bypass security, consent, access controls, or paywalls;
• perform purchases, legal acceptances, account changes, or other impactful actions without appropriate user control and review;
• access areas the customer or visitor is not authorized to access;
• operate as an unsupervised substitute for sensitive human judgment.

Chat4U may use tool approval defaults, URL allowlisting, deny selectors, and browser-side redaction for common sensitive fields. These measures are risk reducers, not guarantees.

---

6. Prohibited and High-Risk Uses

Customers must not use Chat4U for prohibited purposes or High-Risk Use contrary to the Agreement documents. High-Risk Use includes professional medical, legal, or financial advice, or making or materially informing decisions about employment, housing, credit, insurance, education access, benefits eligibility, essential services, biometric identification or verification, safety-critical matters, government functions, defense activities, or similar high-impact outcomes. High-Risk Use is prohibited unless a separately signed Enterprise agreement expressly authorizes the exact use.

A request for an Enterprise exception may be declined. No plan, configuration setting, or ordinary account approval should be interpreted as a high-risk-use authorization.

---

7. Transparency and Human Reliance

Customers should make clear to end users when they are interacting with AI and should avoid presenting AI output as guaranteed, official, or final for important matters.

Where reliance risk is meaningful, customers should provide escalation to a human support channel and should avoid encouraging users to rely on AI for emergency, professional, or high-impact decisions.

---

8. Safety Monitoring and Remediation

Customers should maintain processes for reviewing assistant performance, correcting knowledge sources, disabling unsafe configurations, and responding to complaints or harmful outputs.

Chat4U may suspend, restrict, or disable accounts, agents, demos, tools, or projects where AI-related behavior creates privacy, security, abuse, legal, or safety risk.

---

9. Changes to This Notice

We may update this notice as product behavior, provider usage, customer safeguards, or legal expectations evolve. The version and last-updated date identify the current notice.

---

10. Contact

Report unsafe AI behavior, prompt injection, unauthorized tool actions, or related concerns to support@chat4u.ai.