These Terms of Service ("Terms") govern your use of Thyra ("Service"), an AI-powered agent platform. By creating an account, signing in, or using the Service, you agree to these Terms. If you do not agree, do not use the Service.
1. Description of Service
Thyra is an AI agent platform that provides:
- AI-powered chat with configurable agents
- RAG(retrieval-augmented generation) knowledge bases
- Integrations with email, calendars, issue tracking, bookings, and other third-party services (see Section 4)
- MCP(Model Context Protocol) servers and extensions
- Voice agents using third-party speech and communications services
- Code execution in isolated sandboxed environments
- Web search via third-party search providers
- Human-in-the-loopescalation for tasks requiring human review
2. EU AI Act Compliance
Thyra is designed with the EU Artificial Intelligence Act (Regulation (EU) 2024/1689) in mind. As a deployer of AI systems, we implement transparency and governance measures consistent with the Act.
2.1 Transparency (Article 50 – Limited-Risk AI)
You are informed that:
- You interact with artificial intelligence when using Thyra agents. Responses are generated by AI models from one or more third-party providers and are not authored by humans unless explicitly indicated.
- AI outputs may be inaccurate, incomplete, or inappropriate. Do not rely on them for critical decisions without verification.
- For AI-generated text published on matters of public interest, we disclose that content is AI-generated unless it has undergone human editorial review with clear editorial responsibility.
2.2 Human Oversight
- Thyra supports human-in-the-loopworkflows. Agents can escalate to human reviewers when configured.
- Certain actions (e.g., placing calls, sending emails in autonomous mode) may require explicit user or human approval.
- You may configure agents to operate in draft mode, where proposed actions are reviewed before execution.
2.3 Prohibited Uses
You must not use Thyra for:
- Prohibited AI practices under the EU AI Act (e.g., subliminal techniques beyond a person's consciousness, exploitation of vulnerabilities, social scoring, real-time biometric identification in publicly accessible spaces for law enforcement, except as permitted by law)
- Illegal activities or violation of applicable laws
- Harm, harassment, or deception of others
- Spam or unsolicited communications via connected services (e.g., email or calendar integrations)
- Circumvention of security, rate limits, or access controls
2.4 Risk Classification
Thyra is intended as a limited-risk AI system under the EU AI Act: it interacts with users and generates synthetic text. We do not deploy high-risk AI systems (e.g.,
3. Account & Registration
- You must provide accurate information when registering.
- You are responsible for maintaining the confidentiality of your password and for all activity under your account.
- Sessions use secure cookies for authentication.
- API tokens may be created for programmatic access; you are responsible for securing them.
4. Integrations & Third-Party Services
When you connect third-party services, you authorize us to access and use your data as needed to provide the Service. Your use of those services remains subject to their respective terms and privacy policies. We are not responsible for the availability, accuracy, or conduct of third-party services.
4.1 Email integrations
By connecting an email account, you authorize us to read, send, reply, and (where supported) forward emails and to poll for new messages when connected. You are responsible for ensuring that your use of email through the Service complies with the provider’s policies and anti-spam laws.
4.2 Calendar integrations
By connecting a calendar, you authorize us to access and, where supported, create, update, or delete events. Calendar data may be used to provide context to AI agents. You may configure privacy options (e.g., hiding private event details).
5. Extensions, MCP Servers & Tools
You may attach MCP servers and other extensions to agents. These may execute custom logic and access external systems. You are responsible for the configuration and use of such extensions. We do not warrant the security or correctness of third-party MCP servers or extensions.
6. AI & LLM Usage
- Responses are generated by AI models from one or more third-party providers. We do not guarantee accuracy, completeness, or suitability for any purpose.
- Do not rely on AI output for critical decisions without verification.
- Your inputs and context may be sent to these providers as described in our Privacy Policy.
7. Voice & Communications
Voice features use third-party speech-to-text and text-to-speech services. Call audio and metadata are processed as described in our Privacy Policy. You must comply with applicable telecommunications and consent requirements when using voice features.
8. Code Execution
Agents may run code in isolated sandboxed environments. Code execution is subject to resource limits, timeouts, and network restrictions. You are responsible for the code generated or provided by agents. We do not guarantee that code execution is error-free or secure against all attack vectors.
9. Web Search
When the Web Search tool is enabled, queries may be sent to third-party search providers. Search results are incorporated into AI responses. We do not control the content of third-party search results.
10. Credits & Billing
Usage may be tracked in credits (e.g., token consumption). Billing terms, if any, are set forth in separate agreements or pricing pages. You are responsible for any charges incurred under your account.
11. Intellectual Property
We retain ownership of the Service, including software, design, and branding. You retain ownership of your content (e.g., documents, prompts, messages). By using the Service, you grant us a license to process your content as necessary to provide the Service.
12. Disclaimers
THE SERVICE IS PROVIDED "AS IS" AND "AS AVAILABLE." WE DISCLAIM ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, AND NON-INFRINGEMENT. WE DO NOT GUARANTEE UPTIME, ACCURACY, OR UNINTERRUPTED ACCESS. AI OUTPUTS ARE NOT GUARANTEED TO BE ACCURATE, COMPLETE, OR APPROPRIATE.
13. Limitation of Liability
TO THE MAXIMUM EXTENT PERMITTED BY LAW, WE SHALL NOT BE LIABLE FOR ANY INDIRECT, INCIDENTAL, SPECIAL, CONSEQUENTIAL, OR PUNITIVE DAMAGES, OR FOR LOSS OF PROFITS, DATA, OR GOODWILL, ARISING FROM YOUR USE OF THE SERVICE. OUR TOTAL LIABILITY SHALL NOT EXCEED THE AMOUNT YOU PAID US IN THE TWELVE (12) MONTHS PRECEDING THE CLAIM.
14. Termination
We may suspend or terminate your account at any time for violation of these Terms or for any other reason. You may terminate your account by contacting us. Upon termination, your right to use the Service ceases. We may retain data as required by law or our policies.
15. Changes
We may modify these Terms from time to time. We will notify you of material changes (e.g., by email or in-app notice). Continued use after changes constitutes acceptance. If you do not agree, you must stop using the Service.
16. Governing Law
Thyra operates from Delaware, United States and Portugal. These Terms shall be governed by:
- Users in the European Union (including Portugal): The laws of Portugal and the European Union, as applicable. Nothing in these Terms excludes your statutory rights under applicable consumer protection law.
- Users elsewhere: The laws of the State of Delaware, United States, without regard to conflict of law principles.
17. Contact
For questions about these Terms, contact us at: legal@
Cookies. We use session cookies for authentication and may use analytics cookies where permitted. Cookie use is described in our Cookie Policy. By using the Service, you consent to our use of cookies as set out there.