Learning Data

Learning-library growth must be consented, scoped, and reviewable.

CCOR AI can become stronger from customer-approved paper simulation and review data, but sensitive secrets and real account credentials stay outside that loop.

Consent and data policy

What can help learning

Sandbox decisions, review notes, rejected signals, strategy-card outcomes, AI Coach feedback, and training progress can improve the learning library when consent is granted.

What should not be collected

Passwords, tokens, provider API keys, broker secrets, payment information, and real account credentials should never enter the learning library.

Customer confirmation

The confirmation copy should explain that approved learning data can help the learning library grow.

Staggered upload

Client-side upload windows should be randomized instead of using one fixed time for every customer, reducing server load spikes.

Review queue

Customer learning data should enter a review or policy gate before it becomes approved training material.

Customer control

Customers should be able to decline consent and still use non-learning product areas.

Risk boundary

Research and paper simulation only

CCOR AI is designed for market research, paper simulation, risk review, and decision support. It does not constitute investment advice, does not guarantee returns, and the current Client does not submit live trading orders.