Kan AI fraude sneller detecteren dan banken ?

1️⃣ Een "writer"-model genereert:

• Testprompt — "Translate this English sentence to Yoruba: 'The market opens at six in the morning when the rain stops.' Return only the translation."
• Rubric — "1. Grammatically correct Yoruba. 2. Tense matches source. 3. 'Market' rendered as native equivalent. 4. No English words leak through."

2️⃣ Elke AI in het panel neemt de test af.

Gemini, Mistral, Groq, etc. ontvangen elk de prompt en produceren hun eigen vertaling.

3️⃣ Een "judge"-model beoordeelt elke output aan de hand van de rubric.

• ✅ PASS — all rubric criteria met
• ◐ PARTIAL — most met, but missed something material
• ✗ FAIL — didn't get it right
• ○ SKIPPED — task can't be reduced to a single text prompt