After reading about the “suicide” of yet another whistleblower, it got me thinking.
When working at large enough company, it’s entirely possible that at some point you will get across some information the company does not want to be made public, but your ethics mandate you blow the whistle. So, I was wondering if I were in that position how I would approach creating a dead man’s switch in order to protect myself.
From wikipedia:
A dead man’s switch is a switch that is designed to be activated or deactivated if the human operator becomes incapacitated, such as through death, loss of consciousness, or being bodily removed from control. Originally applied to switches on a vehicle or machine, it has since come to be used to describe other intangible uses, as in computer software.
In this context, a dead man’s switch would trigger the release of information. Some additional requirements could include:
- No single point of failure. (aka a usb can be stolen, your family can be killed, etc)
- Make the existence of the switch public. (aka make sure people know of your mutually assured destruction)
- Secrets should be safe until you die, disappear, or otherwise choose to make them public.
Anyway, how would you go about it?
Making the existence of the switch public is often something you don’t want. It allows others to do troubleshooting in advance. It also destroys your reputation with many people who might otherwise work with you.
If you are content to keep things secret, share the documents with several different friends or law firms in several different countries along with conditions for release. Don’t tell them or everyone who all has the documents. That sounds relatively simple.
I agree with all of the above, except I’d add encryption to the data.
That way you are not putting your life in their hands, at least until it doesn’t matter / you want the data released. Encryption keys are super lightweight vs data; taken to an unreasonable extreme, a KB could unlock TBs.
Though you’d probably want something more like a passphrase. Anyway, that basic idea is sound but I dunno about the exact delivery/delay mechanism. Gun to my head and I have seconds to decide… scheduled send from a major cloud email provider, pay way in advance, and an increasing flood of calendar events/reminders up to the day it sends. The message would include enough information about the encryption used and formats within that any tier 1 helpdesk level IT person could access the data.
Not perfect, but a good enough balance of simple and robust to start with.