While I respect Snowden and find many of the claims plausible much of the evidence and story didn’t add up. Here are a few doubts I have, and I’d like to see the rebuttal.
First, he seemed to embellish his role and authority at NSA. Based on his vernacular, and the artifacts that he presented, I believe he was an IT admin, who happened upon these files. His archive was predominantly design / spec docs with screenshots, rather than code artifacts (compiled code, 1st hand screenshots, configs). In his interviews he didn’t seem to have a deep technical grasp of the systems, and he was cagey about his areas of responsibility.
Second, XKeyScore and other system implementations seemed too primitive for the claimed capability. It was claimed they could google everyone’s email (or some large share of the population, let’s say 100-500m individuals, in real time), but the designs were primitive collegiate level , single-instance mySQL databases and CRUD UIs. I didn’t see any indications the system capability met the claims .
I’m not an expert, I’ve read No Place to Hide, watched some interviews, and used to follow him on Twitter.
I’d like to see how this community assessed him personally and the artifacts he shared.
* and I don’t remember Snowden presenting evidence that NSA (et al) had the datacenter / compute capacity to surveil these targets. Even a fraction of the target traffic (3+ wireless carriers, dozens of messaging email software firms) would require a fraction of their datacenter capacity to sniff, trace, index, analyze . Was that capacity presented in the archive?
He was a sysadmin who built a system that gathered the documents. Why do you think that SQL is not capable of searching email? Also an anonymizing system already existed before 2001. The safeguards were dropped afterwards.