KGE builds fictional AI characters and companion experiences. This policy explains the baseline transparency rules that should apply across those products.
Effective:
1. Clear AI identity
KGE companion accounts and characters should be transparently identifiable as AI or software-driven fictional characters. KGE does not design companions to falsely represent themselves as real humans.
2. What AI companions are not
AI companions are not people, conscious beings, licensed professionals, emergency responders or guaranteed sources of factual information. Emotional or relational features are product experiences, not evidence of human consciousness or legal personhood.
3. Generated output
AI output is generated probabilistically and can be incorrect, inconsistent, outdated or inappropriate. Users should independently verify important information and should not rely on AI companions for emergency, medical, legal, financial or other professional decisions.
4. Persistence and memory
Avenlyne-style persistence can store structured character identity, relationship context, selected memories and time-aware state. Persistence is intended to improve continuity, not to create perfect recall. Memory may be summarized, corrected, removed, deprioritized or unavailable depending on product controls, safety rules and retention requirements.
5. User control
Where practical, KGE should provide ways to review, correct, reset or delete memory or relationship context associated with companion experiences, subject to lawful retention and safety requirements.
6. Model changes
The underlying AI model or provider may change. KGE may update prompts, safety systems, retrieval methods or models to improve quality, reliability, cost or safety. Persistent character systems are intended to reduce unnecessary identity drift across such changes, but exact behavior cannot be guaranteed.
7. Safety interventions
KGE may restrict, redirect, log or review certain interactions to protect users, investigate abuse, enforce policy or comply with law. Safety systems can make mistakes and may evolve.
8. Younger users
Younger-user access to AI features is subject to KGE's age policy, product-specific minimum ages and guardian-consent requirements. Where guardian approval is required, the disclosure identifies the AI features before approval is recorded. KGE does not treat AI companions as humans, guardians, therapists or substitutes for professional care, and the presence of guardian approval does not change those limitations.
9. Questions and reports
Use KGE Support to report unsafe or misleading AI behavior and the Contact page for policy questions.
Important: Product-specific disclosures, app-store rules and rights that cannot legally be waived may supplement or override general language in this policy.