The answer to the literal question of whether one can prevent re-encryption would be no.
Presumably your goal is to prevent the customer from tampering with the data, i.e. altering it and misrepresenting it as the original. You also want to avoid actually keeping a decryptable copy of the data.
Since you need only to encrypt and the customer only to decrypt, the straightforward solution is public key cryptography and authentication. The customer would send you their public key, which you would use to encrypt the data. You would also sign the data with your private signing key, and share your public verification key to prove that you've signed it.
The customer can then decrypt the data with their private key, which you don't have. But while they can encrypt another piece of data, they can't sign it, as they don't have your private signing keys.
With this solution, you can actually delete both your plaintext and your ciphertext. Your public verification key will suffice to conclusively prove whether any given message has or hasn't been signed by you.