What has been found to undermine the public key infrastructure system?

The entire Public Key Infrastructure (PKI) model is built on a chain of trust that originates from a trusted third party, the Certificate Authority (CA). If a CA is compromised or acts maliciously (i.e., becomes disreputable), it can issue fraudulent certificates that legitimate clients (like web browsers) will trust. This fundamentally breaks the assurance that a public key belongs to the claimed entity, allowing for successful man-in-the-middle attacks and undermining the entire system's integrity. Historical breaches at CAs like DigiNotar have demonstrated this systemic risk.

A. PKI is designed to prevent man-in-the-middle attacks; such attacks are a consequence of a compromised PKI, not the cause of its undermining.
B. Mechanisms like Certificate Revocation Lists (CRLs) and OCSP exist specifically to handle compromised or abandoned keys.
D. Browsers not having a CA's key is a feature of the trust model; it means the CA is untrusted, which is how the system is designed to work.

1. Holz, T., Sasse, L., & Teufl, P. (2012). Taming the Public Key Infrastructure. In Availability, Reliability, and Security in Information Systems and HCI (pp. 1-15). Springer. Section 1 discusses the "single point of failure" nature of CAs. https://doi.org/10.1007/978-3-642-32498-71

2. Clark, P. C., & van Oorschot, P. C. (2013). SoK: The Broken Triangle: A Review of Faults in PKI. In 2013 IEEE Symposium on Security and Privacy (pp. 553-567). Section III.A discusses CA compromise as a primary threat. https://doi.org/10.1109/SP.2013.43

3. Cooper, D., et al. (2010). Internet X.509 Public Key Infrastructure Certificate and Certificate Revocation List (CRL) Profile (RFC 5280). IETF. Section 1.1 outlines the trust model dependent on CAs.