EVT 2026-02-19 - Stage A2 AI Entitlement Clarity and Credit Fairness

Session scope

Finalize execution-level rules for AI access boundaries by tier and credit fairness behavior in result flows.

Confirmed outcomes

1. Free result visibility for Writing/Speaking:

• Show only basic total score.
• Do not open AI detail analysis blocks.

2. Pro AI detail policy:

• Pro remains basic-scoring only for Writing/Speaking.
• Detailed AI analysis stays locked (advanced AI belongs to Pro Max credit scope).

3. Reopen behavior for already-scored result:

• Reopening the same scored result does not consume additional AI credit.
• Charge-once semantics applies per scoring job/result.

4. AI scoring system failure behavior:

• Auto-refund consumed AI credit when scoring fails due to system error.
• User should not need manual ticket to get refund in normal failure path.

5. Mid-attempt entitlement/credit drop behavior:

• Do not block submission.
• Allow submit as normal, then lock unavailable AI sections at result.
• Show upgrade/top-up suggestion from locked sections.

Open items carried forward

• None in this entitlement-clarity cluster.