Acceptable Use and AI Transparency Policy

Last updated: 30 October 2026

This policy forms part of the Terms of Service. Breach may lead to suspension or termination of the account.

1. Prohibited uses

You may not use Spen to:

Break the law or infringe third-party rights (intellectual property, privacy, reputation, trade secrets).
Send spam, unsolicited commercial communications in breach of applicable law, or bulk messages that violate Channel policies.
Impersonate another person or mislead recipients about who is sending a message.
Harass, threaten or discriminate, or distribute hateful or violent content or sexual content involving minors.
Commit fraud, phishing, scams, money laundering or sanctions evasion.
Monitor or profile people without a legal basis, or process special categories of data (health, ethnic origin, beliefs, etc.) without complying with the GDPR.
Engage in AI practices prohibited by Article 5 of Regulation (EU) 2024/1689 (AI Act), such as subliminal manipulation, exploitation of vulnerabilities or social scoring.
Make automated decisions with legal or similarly significant effects on people (hiring, credit, insurance, access to essential services) without human oversight and without complying with applicable law.
Develop weapons or malware, or attack computer systems.
Circumvent limits, security measures or controls of Spen or the Integrations, reverse engineer Spen, or extract its models, prompts or data.
Resell or give third parties access to Spen without Mazinger's authorization.

Review Outputs before using or sending them, especially for significant decisions.
Configure automated Routines carefully: whatever runs without your approval runs on your instructions.
Only connect accounts and Integrations you are authorized to use.
If you use Spen within an organization, follow its internal policies on AI use and data processing.
Where required, inform recipients that content was generated or assisted by AI.

It is an AI. Spen is an artificial intelligence system. We tell you so at sign-up and at the start of each Channel.
Models. Spen uses several third-party AI models (including from Anthropic, OpenAI, Google, MiniMax and open models) and selects the most suitable one for each task. In private cloud and self-hosted deployments, the Customer may choose its own models.
Human oversight. By default, Spen does not send communications or create or modify information without your approval. You can see what Spen has done and reverse its actions where the connected tool allows.
Limitations. AI models can make mistakes, invent information or misread instructions. Spen does not guarantee the accuracy of Outputs.
Data. Spen does not train models on your information. See the Privacy Policy.

If you detect misuse of Spen or receive a message sent with Spen that you consider abusive, write to abuse@spen.ai.