You notice the contents of the message appear to be encoded in some manner to defy casual inspection. You decide to take pictures of the package and its contents with your phone, then return the items to the place you found them.
Returning home you eagerly examine your photographs. Because you're clever you eventually decode the messages captured in your pictures. Apparently a foreign intelligence service (FIS) is using the dead drop to communicate with spies in your area! You're able to determine the identities of several Americans working for the FIS, as well as the identities of their FIS handlers. You can't believe it. What should you do?
You decide to take this information to the world via your blog. You found the messages on your own, and you did the work to understand what they mean. If the press reads about your discovery, they'll likely take it farther.
You consider going to the press first, but you decide that it won't hurt to drive traffic to your own blog first. You might even be able to launch that small private investigator practice you've always wanted!
After publishing your post, the press indeed notices, and publishes an expose featuring an interview with you. Several US intelligence agencies also notice. They had been monitoring the dead drop themselves for a year, and had been working a complex joint case against all of the parties you identified. Now all of that work is ruined.
Before the intelligence agencies can react to your disclosure, the targets of their investigation disappear. They will likely be replaced by other agents quickly enough, using other modes of communication unknown to the US agencies. The FIS will alter their operation to account for the disclosure, but it will continue in some form.
That is the problem with irresponsible disclosure. To apply the situation to the digital security world, make the following changes.
- Substitute "command and control server" for "dead drop."
- Substitute "tools, exploits, and other digital artifacts" for "messages."
- When the adversary learns of the disclosure, they move to other C2 infrastructure and develop or adopt new tools, tactics, and procedures (TTPs).
What should the hypothetical "security researcher" have done in this case?
It's fairly obvious he should have approached the FBI himself. They would have realized that he had stumbled upon an active investigation, and counseled him to stay quiet for the sake of national security.
What should "security researchers" in the digital world do?
This has been an active topic in a private mailing list in which I participate. We've been frustrated by what many of us consider to be "irresponsible disclosures." We agree that sharing threat intelligence is valuable, but we prefer to keep the information within channels among peers trusted to not alert the adversary to our knowledge of intruder TTPs.
Granted, this is a difficult line to walk, as I Tweeted yesterday:
Responsible security intel teams walk a fine line between sharing for the benefit of peers and risking disclosure to the detriment of all.
The best I can say at this point is to keep this story in mind the next time you stumble upon a package in the woods. The adversary is watching.