One of the possible attacks that we would want to protect our model against is MITM attacks. A MITM attack occurs when an intruder tries to eavesdrop on a supposedly private communication to deploy a trained machine learning model.

Let's try to understand MITM attacks sequentially using an example scenario.

Let's assume that Bob and Alice want to exchange messages using PKI:

1. Bob is using {Pr_Bob, Pu_Bob} and Alice is using {Pr_Alice, Pu_Alice}. Bob has created a message, M_Bob, and Alice has created a message, M_Alice. They want to exchange these messages with each other in a secure way.
2. Initially, they need to exchange their public keys to establish a secure connection with each other. This means that Bob uses Pu_Alice to encrypt M_Bob before sending the message to Alice. 
3. Let's assume that we have an eavesdropper, X, who is using {Pr_X, Pu_X}. The attacker is able to intercept the public key exchanges between Bob and Alice and replace them with its own public certificate. 
4. Bob sends M_Bob to Alice, encrypting it with Pu_X instead of Pu_Alice, wrongfully thinking that this is Alice's public certificate. Eavesdropper X intercepts the communication. It intercepts the M_Bob message and decrypts it using Pr_Bob.

This MITM attack is shown in the following diagram:

Now, let's look at how we can prevent MITM attacks.