Intelligence,
with accountability.
Ciel Asylum uses artificial intelligence as a tool for acceleration, exploration, and automation, not as a substitute for human direction, professional judgment, or responsibility for the work we deliver.
Useful technology. Clear boundaries.
AI is most valuable when its role is understood. These principles guide how Ciel Asylum approaches AI-assisted work, client solutions, and the AI experiences we build.
People remain responsible.
Strategy, creative direction, material decisions, review, and client accountability remain human-led even when AI contributes to the process.
AI should solve something.
We use AI where it can improve speed, exploration, access to information, or workflow, not simply because a tool is available.
Share only what is needed.
We aim to limit personal, proprietary, or confidential information supplied to AI systems to what is reasonably necessary for the task.
Material AI use should not be hidden.
When AI is a meaningful part of a client-facing system, experience, or deliverable, we aim to explain its role in a way that is useful to the people affected.
1. Purpose of This Statement
This AI Transparency statement explains how Ciel Asylum may use artificial intelligence and automation in its own operations, professional services, creative work, technical work, and website experiences.
AI tools and provider capabilities change quickly. This statement describes Ciel Asylum's current principles and intended practices rather than promising that every AI tool, model, or workflow will operate in exactly the same way.
2. How Ciel Asylum May Use AI
Ciel Asylum may use AI when requested, appropriate, or useful to support work such as:
The role of AI can range from minor behind-the-scenes assistance to a material part of a client-facing product or workflow. The level of review, disclosure, testing, and governance should increase with the significance and risk of the use case.
3. Human Direction & Review
Ciel Asylum does not treat AI output as automatically correct, complete, original, secure, or ready for publication. Human direction and review remain important parts of our process.
Depending on the engagement, human review may include checking factual claims, editing language, evaluating design quality, reviewing code, testing workflows, considering privacy and security, and assessing whether an AI-assisted result is appropriate for its intended use.
AI may accelerate execution. It does not transfer responsibility away from Ciel Asylum or the human decision-makers involved in the project.
4. Disclosure & Labeling
Ciel Asylum aims to disclose AI involvement when that involvement is material to understanding a client-facing system, experience, deliverable, or interaction.
Routine productivity assistance, such as brainstorming, drafting alternatives, code assistance, summarization, or internal research, may not be individually labeled when AI is not itself a material feature of the final experience.
AI-generated or synthetic media may be identified when disclosure is reasonably necessary to avoid misleading a viewer or user about what they are seeing, hearing, or interacting with.
When Ciel Asylum builds an AI-powered feature for a client, the appropriate disclosure approach should be determined by the use case, audience, applicable law, platform requirements, and the degree to which users may reasonably believe they are interacting with a human.
5. Client Data, Privacy & Confidentiality
Ciel Asylum aims to limit personal, proprietary, or confidential information supplied to AI systems to what is reasonably necessary for the task and to use appropriate provider settings, account configurations, contractual controls, or alternative workflows where available and appropriate.
AI providers may process prompts, files, outputs, metadata, or usage information according to their own product terms, privacy practices, enterprise settings, and technical architecture. Where a project involves sensitive data or heightened confidentiality requirements, the AI workflow should be evaluated specifically for that engagement rather than assumed to be suitable.
Additional information about personal information is available in the Privacy Policy.
6. Ciel Asylum Website AI Assistant
Ciel Asylum intends to deploy an AI-assisted experience on its website. When introduced, the assistant should be presented as an AI system rather than as a human employee.
The assistant may be used to answer approved questions, guide visitors, help identify services, collect project context, support navigation, or connect visitors with other Ciel Asylum workflows.
AI assistant responses may be incomplete, inaccurate, or based on limited context. Visitors should not rely on the assistant as a source of legal, medical, financial, regulatory, or other licensed professional advice.
Visitors should avoid submitting unnecessary sensitive or confidential information through an AI assistant. Where appropriate, the assistant may direct a visitor to a human contact or another Ciel Asylum channel.
7. Accuracy, Reliability & Limitations
Generative AI can produce inaccurate statements, fabricated details, outdated information, insecure or defective code, unexpected visual similarities, inconsistent reasoning, or answers that sound confident despite uncertainty.
Ciel Asylum therefore treats AI output as material that may require verification, testing, editing, or rejection depending on the context. Higher-risk uses require greater caution than low-risk creative exploration or routine productivity assistance.
Ciel Asylum does not guarantee that an AI-enabled system will always produce a particular answer, remain available without interruption, or behave identically after a third-party provider updates its model, service, safeguards, or policies.
8. Bias, Fairness & Sensitive Use Cases
AI systems can reflect limitations or biases in training data, model design, evaluation methods, prompts, workflows, or human decisions around their use. Ciel Asylum aims to consider whether outputs or automated processes could create unfair, discriminatory, or misleading effects.
AI used in sensitive or consequential contexts, such as employment, housing, lending, healthcare, education, eligibility, or other regulated decision-making, requires case-specific legal, technical, and governance review beyond a general creative or automation workflow.
Ciel Asylum may decline, limit, redesign, or require additional safeguards for an AI use case when the risks are not appropriate for the proposed implementation.
9. AI-Assisted Creative Work & Intellectual Property
AI-assisted creative work can raise questions involving authorship, originality, licensing, training data, trademarks, publicity rights, and similarity to existing material. Ciel Asylum does not assume that every AI-generated output can be owned, registered, or used without restriction.
Our process may include human creative direction, selection, editing, compositing, rewriting, coding, or other development around AI-assisted material. Project-specific rights, licenses, and ownership are governed by the applicable project agreement and Ciel Asylum's Terms of Service.
Where a project requires specific ownership, exclusivity, registration, provenance, or licensing assurances, those requirements should be identified before relying on an AI-assisted production workflow.
10. Third-Party AI Providers
Ciel Asylum may use third-party AI models, APIs, platforms, automation services, or software. Those providers operate under their own terms, privacy practices, usage limits, safety controls, intellectual-property terms, retention practices, and technical policies.
Provider selection may vary by project because no single AI system is appropriate for every task. Ciel Asylum may consider factors such as capability, privacy, security, reliability, integration requirements, cost, contractual controls, and client needs.
Because third-party AI services evolve rapidly, Ciel Asylum may change providers or implementation methods when doing so improves the project, reduces risk, or responds to provider changes.
11. Governance, Feedback & Updates
Ciel Asylum's AI practices will continue to evolve as tools, standards, regulations, risks, and client expectations change. We may update internal workflows, disclosure practices, review procedures, provider choices, and this transparency statement accordingly.
Questions about AI-assisted services can be sent to support@cielasylum.com. Privacy or data-handling questions can be sent to privacy@cielasylum.com.
The effective date shown on this page should be updated whenever material changes are made to this statement.
Ask what the system does.
And what it doesn’t.
If you want to understand whether AI may be used in your project, how confidential information will be handled, or what safeguards are appropriate for an AI-enabled solution, ask before implementation.