That the FBI arrested a man suspected of using Tor to host child pornography distribution services further fuelled speculation that perhaps US authorities had launched an attack on Tor.
Some infosec specialists quickly analysed the malware and suggested it was controlled by an entity using IP addresses associated with defence contractor Science Applications International Corporation (SAIC) and/or the NSA. One and one were promptly put together to suggest three elements explaining the Tor takedown:
- The arrest of porn suspect Eric Eoin Marques was but one action in a wider attack on Tor
- The US government, probably the NSA, created weaponised malware to take down Tor
- SAIC and/or the NSA were the source and/or controller of that malware
A couple of days down the track, that theory is looking rocky, as two of the organisations that helped the malware theory to spread have issued a joint post saying their initial analysis of the malware was wrong.
Cryptocloud and Baneki Privacy Labs write that their initial analysis of the IP addresses used by the “torsploit” probably don't have anything to do with SAIC. Cryptocloud's also less-than-certain it's earlier assertion that NSA IP addresses were involved is right.
The post we've linked to above is long, rambling and suggests that even if it is not possible to find an IP address tied directly to the NSA in the Torsploit code, the incident looks an awful lot like the kind of thing the NSA is known to be capable of and interested in.
Edward Snowden's recent revelations make it plain that the NSA is peering into a great many dark places. Tor's status as a likely gateway to much of the “dark web” means attempts to gain more intelligence on just what lies within the onion router seem well within the bounds of possibility.
For now, however, the dots aren't joined. Nor, for what it is worth, is a decent explanation of where Torsploit came from or just how much damage it has done. ®