Logic locking induced fault attacks

Logic locking has been presented in the past as a solution to avoid overproduction or product piracy. All solutions so far assume or rely on the protection of the correct locking key. This paper does not target the break of locking schemes but looks at another highly critical security risk which comes with modified locking keys: The enabling of fault attacks. We will analyze the applicability of the risk based on different logic locking methods, key management techniques, and fault analysis techniques. Additionally, we will demonstrate its threat by a modified template based persistent fault analysis on an AES implementation. There, the fault injection is realized by randomly inserted locking gates and a permanently modified locking key.