I've been wondering about this:

http://www.computersecurityarticles.info/antivirus/sophos/fbi-takes-on-coreflood-botnet-but-is-this-a-step-too-far/


...while on first i'd think it were a good idea but i still wonder if, from a security standpoint, sending commands to botnets can be considered 'safe' in that context - your opinion?

Yes and no.........

Yes to show they can be stopped.
No because you are now escalating the fight once they reprogram the bots.

The yes part is good in that an infected box is infected because no one monitors it or because they are idiots to not have proper security, so they should be remotely fixed.

The no part is because the repair will be found and the result will be a stronger program that can resist being exploited.

Better to find a way to block.

My nickles worth .

Woodmann