← Back to Bridge

AI Transparency, Literacy & Responsible Use

Last updated: 9 March 2026

Ulysses32 puts powerful AI tools in your hands. With that power comes responsibility. This page explains how AI works within Ulysses32, what you should understand before using it, and guidelines for responsible use.

1. What Ulysses32 Does (and Doesn’t Do)

Ulysses32 is an orchestration platform. It coordinates and manages AI agents — it does not create the AI itself. Think of it as the bridge of a starship: the bridge gives you controls, screens, and communication channels, but the engines (AI models) are built by others.

  • We provide: The interface, orchestration logic, crew management, integration framework, and user experience.
  • We do not provide: The AI models, training data, or the intelligence itself.
  • AI providers provide: The actual AI models (Claude, GPT-4, Gemini, Mistral, etc.) with their own capabilities, limitations, and policies.

2. How AI Works in Ulysses32

When you interact with AI crew members in Ulysses32:

  1. You send a message or instruction through the Ulysses32 interface.
  2. Ulysses32 formats and routes that instruction to your chosen AI provider’s API.
  3. The AI provider processes the request and returns a response.
  4. Ulysses32 displays and, if configured, acts on that response.

At every step, your data travels directly between your installation and the AI provider. Quantum Touch is never in the middle.

3. Understanding AI Limitations

AI models are powerful tools, but they are not infallible. Every captain should understand these limitations:

Hallucinations

AI models can generate information that sounds confident and authoritative but is factually incorrect. This is known as “hallucination.” Always verify important facts, figures, citations, and claims independently.

Bias

AI models are trained on human-generated data and can reflect biases present in that data. Be aware that outputs may contain cultural, gender, racial, or other biases. Review outputs critically, especially for content intended for public consumption.

Context Limitations

AI models have limited context windows and may lose track of earlier parts of long conversations. They do not have true memory or understanding — they predict likely next tokens based on patterns in training data.

No Real-Time Knowledge

AI models have knowledge cutoff dates and do not have access to real-time information unless explicitly connected to external tools or search capabilities.

4. Your Responsibilities as Captain

As the captain of your Ulysses32 installation, you bear full responsibility for:

  • Reviewing all AI outputs before publishing, sharing, or acting on them.
  • Verifying accuracy of any factual claims, especially in professional, legal, medical, or financial contexts.
  • Ensuring compliance with the terms of service of your chosen AI providers.
  • Complying with applicable laws in your jurisdiction regarding AI-generated content, including disclosure requirements.
  • Disclosing AI involvement where required by law or professional standards.
  • Protecting sensitive data — be mindful of what you send to AI providers.

5. Safe Installation Practices

AI behaviour is inherently unpredictable. Crew members may misinterpret instructions, take unintended actions through integrations, or produce unexpected outputs. To protect your systems and data:

  • Use a sandboxed environment — Docker containers, VMs, or dedicated servers are strongly recommended.
  • Apply least privilege — only grant the permissions your crew actually needs.
  • Isolate from critical systems — don’t install alongside production databases, financial systems, or sensitive infrastructure.
  • Maintain backups — keep independent backups of anything important on the host system.
  • Review integrations carefully — before connecting tools, understand what access they grant to AI crew members.

Running Ulysses32 outside of an isolated environment is at your own risk. See our Terms of Service (Section 9) for full details.

6. Responsible Use Guidelines

We ask all Ulysses32 users to commit to responsible AI use:

Do

  • Use AI as a tool to augment your capabilities, not replace critical thinking.
  • Verify important information generated by AI before acting on it.
  • Disclose the use of AI-generated content where transparency is expected or required.
  • Review your AI provider’s terms of service and acceptable use policies.
  • Keep your AI provider API keys secure and never share them.
  • Report bugs or unexpected AI behaviour to both us and the relevant AI provider.

Don’t

  • Generate content that is illegal, harmful, threatening, abusive, or violates the rights of others.
  • Use AI to create deepfakes, misinformation, or deceptive content intended to mislead.
  • Rely on AI for critical decisions without independent verification (medical, legal, financial, safety).
  • Attempt to circumvent AI provider safety measures or content policies.
  • Use AI to impersonate real individuals without their explicit consent.
  • Send highly sensitive personal data (government IDs, medical records, financial details) to AI providers unless you have reviewed and accepted their data handling practices.

7. AI Provider Terms

Each AI provider has its own terms of service, acceptable use policies, and content policies. As captain, it is your responsibility to review and comply with these. Key providers commonly used with Ulysses32:

8. Platform Intellectual Property

While your AI-generated outputs are yours, the Ulysses32 platform itself — including its source code, interface design, orchestration logic, and all associated materials — is the proprietary intellectual property of Quantum Touch Limited.

Your right to use the Software does not include any right to copy, extract, modify, redistribute, or create derivative works from the platform’s code or design. The use of open-source components within Ulysses32 does not make the platform open-source. See our Terms of Service (Sections 4, 5, and 6) for full intellectual property terms.

In the context of AI transparency: the AI models generate content for you, but the tool you use to orchestrate them (Ulysses32) is protected. You own the paintings — you don’t own the brushes.

9. Emerging Regulation

AI regulation is evolving rapidly worldwide. Key frameworks include:

  • EU AI Act — Classifies AI systems by risk level and imposes obligations accordingly. As an orchestration tool for general-purpose AI, Ulysses32 itself is not classified as a high-risk AI system.
  • GDPR — Applies when personal data is processed. As the data controller for your installation, you are responsible for GDPR compliance regarding data you send to AI providers.
  • National regulations — Various countries are introducing AI-specific laws. Stay informed about regulations in your jurisdiction.

10. Our Commitment

Quantum Touch is committed to:

  • Transparency — Being clear about what Ulysses32 does and doesn’t do.
  • Privacy by design — Building software that respects your data sovereignty.
  • User empowerment — Giving you the tools and information to use AI responsibly.
  • Ongoing updates — Keeping this page current as AI technology and regulation evolve.

11. Contact

Questions about AI transparency or responsible use? legal@strangematerials.com

Quantum Touch Limited
Dublin, Ireland