The secure sockets layer (SSL) credentials were digitally signed by a valid certificate authority, an imprimatur that caused most mainstream browsers to place an HTTPS in front of the addresses and display other logos certifying that the connection was the one authorized by Google. In fact, the certificates were unauthorized duplicates that were issued in violation of rules established by browser manufacturers and certificate authority services.
The certificates were issued by an intermediate certificate authority linked to the Agence nationale de la sécurité des systèmes d’information, the French cyberdefense agency better known as ANSSI. After Google brought the certificates to the attention of agency officials, the officials said the intermediate certificate was used in a commercial device on a private network to inspect encrypted traffic with the knowledge of end users, Google security engineer Adam Langley wrote in a blog post published over the weekend. Google updated its Chrome browser to reject all certificates signed by the intermediate authority and asked other browser makers to do the same. Firefox developer Mozilla and Microsoft, developer of Internet Explorer have followed suit. ANSSI later blamed the mistake on human error. It said it had no security consequences for the French administration or the general public, but the agency has revoked the certificate anyway.
An intermediate certificate authority is a crucial link in the “chain of trust” that’s key in connections protected by SSL and its successor protocol, known as transport layer security (TLS). Because intermediate certificates are signed by a root certificate embedded in the browser, they have the ability to mint an unlimited number of digital certificates for virtually any site. The individual certificates will be accepted by default by most browsers.
Maybe it’s time to dust off and update my article on digital signatures and digital certificates, The Essential Role of Trusted Third Parties in Electronic Commerce, 75 Ore. L. Rev. 49 (1996). I think this was the first article published in a US law review on the topic, and even though it’s held up well, there have been many developments in nearly 20 years. On the other hand, there are three new papers I need to finish first…