AI & Translation Policy

Version 1 · effective 2026-08-09 · last updated 8/9/2026

Meaning, not word-for-word

E Klear uses AI to interpret what you actually mean, including slang and context, rather than swapping words one by one.

Translations can be imperfect

Sometimes a translation will miss nuance or regional usage. Treat output as a helpful draft, not a certified translation.

Uncertainty is shown, not hidden

Where the system is not confident about an expression, it says so instead of inventing a meaning.

Malformed AI responses are rejected

Every AI response is validated server-side against a strict structure. Invalid responses are retried once with a stricter prompt, and fall back safely if still invalid.

Report and correct

You can correct a translation result. Corrections are treated as learning signals for review — they never instantly rewrite the shared dictionary.