FAIR USE POLICY AND AI CREDIT PROCESSING
Purpose and Scope
The present Fair Use Policy (hereinafter, the "Policy") governs the allocation, consumption and monitoring of Artificial Intelligence computing capacity (tokens and requests to integrated APIs) for the AligentGrow Agent Suite (Mail Pilot, Meeting Strategist, Smart Talent, KPI Performance, Innovation AI and Brand AI).
This Policy aims to guarantee the ultra-speed and continuity of the infrastructure based on Google's Gemini models, maintain a predictable cost structure and high return on investment (ROI) for the client in the Colombian and Latin American market, and ensure compliance with regulations on truthful advertising (Law 1480 of 2011), data protection (Law 1581 of 2012) and governance of high-risk AI.
Execution Model
- Account Installation: Each agent is installed and integrated autonomously into the Google Workspace mailbox or account corresponding to the role responsible for the process within the company (e.g., the account of seleccion@empresa.com for Smart Talent, or ventas@empresa.com for Mail Pilot).
- Batch processing: The tasks of reading, classifying, and generating documents are executed through periodic automatic triggers (every 5 to 15 minutes), respecting the operational boundaries and execution time limits established in Google Workspace.
- Licensing by Organization: The company acquires a corporate exchange shared by agents, eliminating the costly per-user/seat licensing scheme traditional to the SaaS market.
Operational Units of Measurement
To facilitate the commercial administration of the service, the following equivalencies are established:
- AI Credit (Token): Basic unit of computational processing. In Spanish, 1 token is equivalent to ~4 characters; 1,000 tokens represent approximately 750 words.
- Request / Operation (API Call): Any irreversible task or effect performed by an agent (e.g., rating a CV, drafting an email, extracting minutes from Google Meet, or analyzing a brand manual).
- Monthly Corporate Exchange: Maximum limit of tokens and tasks assigned to the organization's domain according to the contracted plan.
Plan Structure
To safeguard node stability and protect service profitability, AligentGrow employs a system of Double Security Lock (Consumption stops or adjusts upon reaching the Token or Operation limit, whichever comes first):
*Billing for Colombia: Amounts expressed in USD will be settled in Colombian pesos (COP) using the official Representative Market Rate (TRM) on the billing date.
Special Clause for Visual Generation (Brand AI)
Given that AI image generation has a significantly higher atomic computational cost compared to text (up to ~30 times higher per operation):
- Included Capacity: The Start-Up Plan allocates up to 15 images/month, and the Business Plan assigns a limit of up to 50 images/month produced by Brand AI.
- Visual Surplus Packages: Organizations requiring higher graphic volumes can purchase additional image generation packages without altering the fixed word processing fee.
Monitoring, Continuity and Loop Prevention
- Preventive Consumption Alerts: The system issues automatic status alerts to the system administrator when reaching 80% and 95% of the assigned limit.
- Service Continuity (No Abrupt Interruption): Upon reaching 100% of the monthly quota, critical operations are not immediately frozen; a contingency block is enabled while the administrator confirms an upgrade or applies excess charges.
- Excess Fee: Additional usage beyond the contracted data allowance will be billed at infrastructure rates: $0.50 USD per 1,000,000 text tokens and/or $0.01 USD per additional webhook/API operation.
- Protection against Infinite Loops: If the monitoring systems detect a logical failure or infinite loop in the client's letters or emails that generates continuous requests, the execution of the agent will be paused preventively (dead-letter / lock protection), immediately notifying the client's technical contact.
Governance, Legal Transparency and Accountability
In strict alignment with the Colombian Consumer Statute (Law 1480 of 2011), the Personal Data Protection Law (Law 1581 of 2012) and the standards of the European Union AI Law (EU Regulation 2024/1689):
- Mandatory Human Supervision (Human-in-the-Loop):
In selection processes with Smart Talent, AI acts as a support tool by reading CVs, calculating compatibility, and preparing interview questions. The final decision to interview, hire, or reject a candidate ALWAYS rests with a human evaluator.
- Mandatory Human Supervision (Human-in-the-Loop):
- Informative Nature of Analytics:
Advertising investment recommendations (Brand AI) or task creation (Meeting Strategist) are strictly for informational purposes. AligentGrow agents do not execute direct purchases or modify budgets on external advertising platforms independently.
- Third-Party Platform Disclaimer: