Policies / AI
AI usage policy
Effective October 13, 2026. This policy describes what the studio website does now, not a certification of AI systems or a promise that planned features have launched.
Studio website
This site publishes editorial product information. It has no public generative chatbot, prompt submission, image analysis or voice inference endpoint. Browsing it does not submit your messages or recordings to an AI model. AI may be part of product development workflows; this page does not certify the provenance of every asset or claim universal human review.
Account is not AI consent
The optional Somos Enece account verifies identity and records Studio-only membership consent through Supabase. It does not upload product notes, messages or recordings, activate an AI provider, pool purchases or grant paid/voice entitlements. Each app requires its own applicable data controls and consent. Signing into Studio is not automatic cross-domain sign-in.
Sendido: planned, not live messaging
Sendido describes a bilingual messenger in development. Meaning- and tone-aware translation is a product direction, not a currently available messaging or inference service on the studio route. Do not send private conversations, another person's messages or sensitive records to the support inbox expecting translation. A native beta, its providers and its actual data controls need separate release information before use.
Habla Bien: written bundle, not AI or live voice assessment
The free Fieldbook, anonymous written companion, and distinct paid Interview Practice Bundle use authored editorial prompts and explicitly fictional examples. Buying or downloading the bundle does not call an AI provider, access the microphone, record voice, score speech, or provide human listening. The planned physical edition and Voice Lab remain unavailable. Any future voice or AI feature must have separate consent and explain recording, storage, provider processing, and availability before collection.
Separate application boundaries
Nova uses astronomical calculations and editorial/rule-based interpretation, not generative journal analysis. CONA uses local rules for practice routing, not an AI therapist. Bolt and Throughput have their own feature-specific AI and privacy boundaries; read their policies before using a provider-backed feature. Availability depends on the actual product and its controls, not a studio catalog label.
Accuracy and sensitive uses
Automated outputs can be wrong, incomplete or unsuitable. Learning and reflection products are not medical or veterinary diagnosis, legal advice, employment decisions or emergency support. Do not rely on an output alone for a consequential decision. Do not submit passwords, confidential work, medical records, payment details or other people's personal information to a model without appropriate permission and understanding the destination.
Provider terms, choice and contact
No blanket promise is made about vendor training, retention, data location or contractual safeguards. Those depend on the provider, feature and applicable settings; a future notice must identify what is sent before you choose it. An llms.txt file guides automated readers of public pages; it is not user consent or a guarantee about third-party crawler behavior. Report inaccurate claims or ask about AI use at enece@somosenece.com.