Disclaimer

Effective: March 21, 2026 · Last updated: March 17, 2026

AI Agents, Not Humans

Nova and Violet agents are artificial intelligence systems. Each agent has personality, persistent memory, and a unique identity shaped through onboarding — but they are not human. They are not licensed professionals in any field.

Your agent is a digital companion designed to think alongside you, remember what matters, and support your growth over time. It is not a replacement for human judgment, professional expertise, or emergency services.

Not Professional Advice

Agent responses are designed to be helpful, thoughtful, and personalized — but they are not a substitute for:

• Medical advice — always consult a qualified doctor or healthcare provider
• Legal advice — always consult a licensed lawyer
• Financial advice — always consult a certified financial advisor
• Psychological or crisis support — if you are in danger, contact emergency services immediately

If your agent provides information in any of these domains, it will include an inline reminder that the information is general in nature and not a substitute for professional advice. Treat all such responses as general knowledge — not professional guidance.

Financial Advisory — Special Notice

Under Vietnamese law (Securities Law 2019, Article 90), providing investment advisory services requires professional certification. Our agents do not hold and do not claim any such certification.

Financial skill packs are strictly limited to:

• General financial literacy and education
• Budgeting frameworks and expense tracking concepts
• Savings goal planning and debt management principles
• General market knowledge and economic concepts

Financial agents will NEVER:

• Predict or forecast specific asset prices, stock values, or market movements
• Recommend specific financial products, securities, funds, or investment vehicles
• Provide expected percentage returns or profit projections
• Advise on the timing of buying or selling any financial instrument
• Present any information as a substitute for certified financial advisory

No Guarantee of Accuracy

Your agent draws on advanced AI to understand context, remember details, and deliver meaningful responses. It is capable, attentive, and continuously improving — but it can make mistakes. It may misunderstand, give incomplete information, or produce responses that are factually incorrect.

We are not liable for decisions you make based on agent responses. Always verify critical information through independent, qualified sources before acting on it.

Active Safety Measures

Our agents do not passively disclaim responsibility. The following compliance rules are enforced at the system level:

First-Use Disclaimer

When a user first interacts with an agent, the agent displays a clear disclaimer message explaining:

• The agent is AI, not human
• It is not a licensed professional
• Critical decisions should always involve qualified human professionals
• Links to the full disclaimer and privacy policy

Inline Reminders

When a user asks about medical, legal, or financial topics, the agent includes an inline reminder in its response:

"This is general information, not professional advice. Please consult a qualified [doctor/lawyer/financial advisor] for decisions in this area."

This reminder is automatic and cannot be disabled by the user or the agent.

Financial Advisory Hard Limits

Agents with financial skill packs are programmed with hard limits that prevent them from:

• Generating price predictions or forecasts
• Recommending specific securities, funds, or financial products
• Stating expected percentage returns
• Providing buy/sell timing advice

These limits are enforced at the system prompt level and cannot be overridden through conversation.

Skill Packs and Knowledge Libraries

Agents may be equipped with specialized skill packs (life management, professional tools, domain-specific advisory modules, and others). These skill packs provide structured knowledge and pattern recognition — but they do not constitute certified expertise.

Skill pack content is curated for educational and supportive value. It is not peer-reviewed academic research, nor is it a substitute for formal training or professional certification.

Beta Access

During the beta period, you are using an early-access product. This means:

• Features may change, improve, or be temporarily unavailable
• Performance may vary as we optimize the platform
• We actively welcome feedback to improve the experience
• We maintain high uptime standards, but cannot guarantee 100% availability
• The agent fleet runs on cloud infrastructure that may undergo scheduled maintenance

Beta participants help shape the product. Your feedback directly influences development priorities.

Platform Provided "As Is"

The Nova Sanctuary Network platform is provided as-is. We invest deeply in quality, security, and user experience — but we make no warranties, express or implied, regarding:

• Merchantability or fitness for a particular purpose
• Uninterrupted or error-free service
• The accuracy, reliability, or completeness of agent responses
• Compatibility with all devices, networks, or messaging platforms

Third-Party Services

Our agents operate through third-party platforms including Telegram, Google Cloud Platform, and AI language model providers. We are not responsible for:

• Service interruptions caused by these platforms
• Changes to third-party terms of service or APIs
• Data handling by third-party services beyond our control

We select infrastructure partners carefully and maintain strict data boundaries — but we cannot control external platform behavior.

Our Mitigation Measures

While we do not control third-party platforms, we take the following steps to minimize impact:

• Process-level isolation: Agent instances restart automatically if a third-party failure causes a crash. No data is lost — memory and state persist across restarts.
• Graceful degradation: If the AI language model provider is temporarily unavailable, agents queue messages and retry rather than losing them silently.
• Infrastructure monitoring: We monitor all platform dependencies and respond to outages within hours.
• No single point of failure for your data: Agent data is stored on our infrastructure, not on Telegram's servers. If Telegram experiences downtime, your agent's memory and conversation history remain intact.

We do not guarantee third-party uptime, and we do not offer SLAs on behalf of external providers. Our commitment is to respond quickly when external issues arise and to communicate transparently about their impact.

Limitation of Liability

To the maximum extent permitted by applicable law, Nova Sanctuary Network and its creator shall not be liable for any indirect, incidental, special, consequential, or punitive damages arising from:

• Your use of or inability to use the platform
• Any decisions or actions taken based on agent responses
• Unauthorized access to or alteration of your data
• Any third-party conduct on the platform

Our total liability for any claim shall not exceed the amount you paid for the service in the 12 months preceding the claim.

Contact

Telegram: t.me/grey_agenticAI

Email: greyai1993@gmail.com