Protect a Pricing Model Exposed Through a Partner API From Extraction

Your company sells a dynamic pricing model to partner retailers through an API: partners send product and context features (about 25 numeric and categorical fields) and receive a recommended price with a confidence score. Leadership worries that a competitor will sign up as a "partner", call the API many times, and rebuild the model for free.

1. Explain how the attack works and what a competitor needs.
2. What does evidence from a small simulation say about which defenses matter, and what does it say about the defenses people usually reach for first?
3. Propose a layered defense, including detection signals, and discuss what each layer costs the legitimate partners.
4. What would you do about proving theft after the fact?
5. Argue whether the risk is worth the engineering effort.