Back in 2007, Bauer et al. showed that a relay-level adversary could lie about their bandwidth capabilities, thus unfairly increasing the chance of having their relay selected in Tor circuits [^1].
Today, this vulnerability is fixed. How? Do directory authorities verify a relay's advertised bandwidth, and/or do clients? Or has the bandwidth advertisement mechanism been removed such that only directories/clients measure it?
[^1]: Bauer, Kevin, et al. "Low-resource routing attacks against Tor." Proceedings of the 2007 ACM workshop on Privacy in electronic society. ACM, 2007.