← Home

Terms of Use

Effective February 25, 2026

1. General Provisions

These Terms govern the use of the ASI Biont service (hereinafter — "Service"). By using the Service, you agree to these Terms. If you disagree — stop using the Service.

2. Service Description

ASI Biont is an AI-powered platform that writes Python code and executes it in a secure sandbox. The Service provides:

• AI-driven Python code generation and execution in an isolated sandbox
• Universal credential connector for any external service (Gmail, GitHub, GitLab, Telegram, VK, WhatsApp, Discord, Slack, Zoom, Google Drive, Dropbox, Trello, Jira, Notion, PayPal, Stripe, YooKassa, Google Calendar, Figma, WordPress, Twitter/X, Instagram, LinkedIn, YouTube, Docker, AWS, Google Cloud, OpenAI, Shopify, Salesforce, HubSpot and any custom REST API)
• External database queries — AI can execute SQL on your PostgreSQL, MySQL, MongoDB, Redis, ClickHouse, DuckDB, SQLite (AI never sees your connection strings)
• Automatic API discovery — AI loads any REST API's OpenAPI/Swagger spec and understands all endpoints automatically
• OAuth2 authorization — built-in support for Google, GitHub, Slack, Notion OAuth2 flows with auto-refresh
• Webhook receiver — AI accepts webhooks from external services with HMAC-SHA256 verification
• Eternal memory — persistent context across sessions
• Autopilot — scheduled task execution and proactive notifications
• Web search, web scraping, text translation, audio transcription
• File and Git operations (read, write, commit, push)
• Task, goal, note and reminder management
• Integration with Telegram bot (@asibiont_bot) and web dashboard

3. Acceptable Use

The user agrees to:

• Use the Service only for lawful purposes
• Not use AI-generated code for malicious purposes (spam, hacking, fraud, malware)
• Not attempt to bypass the sandbox restrictions or exploit vulnerabilities
• Not violate third-party rights, including intellectual property
• Not use the Service for unauthorized access to systems or data

4. AI Code Generation and Sandbox

AI-generated code runs in a secure sandbox with the following restrictions:

• Only safe builtins are available (bool, int, str, list, dict, len, range, sorted, map, filter, etc.)
• Dangerous modules are blocked: os, subprocess, socket, sys, pathlib, shutil, multiprocessing, ctypes, and others asyncio is available in a safe wrapper (server functions blocked)
• Dynamic code execution (exec, eval, compile, __import__, open, input, breakpoint) is prohibited
• A timeout is enforced on all code execution
• The user is responsible for reviewing and approving AI-generated code before production use

The Service is not liable for damages resulting from AI-generated code, including but not limited to data loss, security breaches, or system failures.

5. Credentials and External Services

The universal credential connector allows the AI to use your credentials for external services:

• Credentials are stored encrypted and used only for API calls on your behalf
• Database connection strings are never exposed to the AI — they are passed server-side only
• OAuth2 tokens are stored per-user and automatically refreshed on expiry
• Webhook payloads are isolated per-user, each event verified with HMAC-SHA256 signature
• You are responsible for the actions performed using your credentials
• The Service does not share your credentials with third parties
• You may revoke access at any time by removing credentials from the Service

6. Disclaimer

The Service is provided "as is". The Operator does not guarantee:

• Uninterrupted and error-free operation of the Service
• Accuracy, completeness, and timeliness of AI-generated code and responses
• That AI-generated code is bug-free, secure, or suitable for a specific purpose
• Availability of specific external services through the universal connector

The user independently reviews and tests AI-generated code before using it. The Operator is not liable for decisions made based on AI recommendations or code output.

7. Limitation of Liability

The Operator is not liable for:

• Losses arising from the use or inability to use the Service
• Damages caused by AI-generated code executed by the user
• Consequences of the user providing inaccurate or compromised credentials
• Temporary unavailability of the Service due to technical reasons
• Actions of third parties who gained access to the user's account

8. Token Policy

• ASI tokens are internal service tokens. 1 token ≈ $0.01 USD.
• Tokens are not securities, investments, CFD, electronic money, or currency.
• Tokens are for Service use only. Purchased tokens are non-refundable, except as required by applicable law.
• 1000 free tokens are provided on signup.
• The Operator may change pricing with user notification.
• Tokens do not grant ownership rights, equity stakes, or voting rights.

9. Account Suspension

The Operator may restrict or block user access in case of:

• Violation of these Terms
• Use of the Service for illegal activities
• Attempts to bypass sandbox restrictions or security systems
• Distribution of malware or spam through the Service

10. Intellectual Property

All rights to the Service, its code, design, and AI models belong to the Operator. Content created by the user (code, tasks, messages) remains the user's property. AI-generated code is provided as a tool output and may be used freely by the user.

11. Changes to Terms

The Operator may amend these Terms. The current version is available at /terms. Continued use of the Service after changes are posted constitutes acceptance.

12. Contact

For questions: @aleksandrinsider.