AI Policy

This document outlines TukuToi’s policy on the use of artificial intelligence (AI). It covers various aspects including its purpose and application, data collection, user consent, transparency, oversight and control, ethics and bias, safety and security, accountability, and continuous updates and maintenance.

In this policy, “TukuToi” means TukuToi LLC, “the Client” means anyone who engages TukuToi’s services, and “users” means anyone who uses tukutoi.com.

Purpose and Application of AI:

TukuToi employs artificial intelligence (AI) in various domains such as programming, content generation, and more. While AI is an integral part of TukuToi’s operations, TukuToi sticks to the Human in the loop (HITL) principle: all decisions and actions facilitated by AI are subject to human oversight and control.

Data Collection, Processing, and Storage:

TukuToi does not store user data beyond what the Privacy Policy describes. Any data sent to third-party AI providers, such as OpenAI, is exclusively for the purpose of obtaining API responses. OpenAI and Anthropic process this data under their data processing terms for business customers and do not use it for training. OpenRouter forwards requests to other model providers whose terms differ, so it is only used where the Client agrees to it.

User Consent:

Users consent to data collection and processing by agreeing to TukuToi’s privacy and usage policy, generally by checking an agreement box. TukuToi respects user autonomy and prioritizes informed consent. Whenever no consent is requested, no User Data is collected for AI interactions.

Transparency:

TukuToi is committed to being transparent with its users. Users will always be informed when they are interacting with an AI system rather than a human.

Oversight and Control:

All AI systems deployed within TukuToi’s operations have human oversight (HITL). TukuToi ensures that AI facilitates but does not replace the human touch in decision-making.

Ethics and Bias:

TukuToi strives for its AI systems to be as unbiased and ethical as possible. TukuToi continuously works towards refining its models and ensuring that they adhere to high ethical standards.

Safety and Security:

To ensure the safety of TukuToi’s AI systems, production data, passwords, keys and personal data are never sent to an AI provider. These measures are taken to prevent malicious attacks and unintended behaviors.

Accountability:

TukuToi is responsible for AI operations and decisions, within the bounds of the Terms and Conditions. TukuToi is not responsible for third party contract breach. TukuToi refuses responsibility over infringement of the providers’ data processing terms by the third parties that are out of its control. No Code, Content or else AI generated data is delivered by TukuToi to Clients without HITL. This website uses AI in one place: messages sent through the contact form are screened for spam by a language model that runs on TukuToi’s own hardware. Nothing is sent to an AI provider, and every message still reaches TukuToi, whatever the verdict; the verdict only decides whether the automatic confirmation email goes out.

Updates and Maintenance:

TukuToi is committed to the continuous improvement and maintenance of its AI systems to serve its users better and ensure the highest quality outcomes.

Third-Party AI Systems:

TukuToi collaborates with third-party AI providers, strictly limited to:

• OpenAI
• Anthropic
• OpenRouter

OpenAI and Anthropic are bound by data processing terms that exclude training on the data sent to them. OpenRouter is only used for a project when the Client agrees to it.