If it gets big enough then people just hear from each other if paying unlocks the data or not.

The best way to end ransomware is to get serious about security. In many cases, being attacked by a ransomware, is paying a low price compared to if it was a targeted attack.

edit: Also, I imagine it gets easier after you wrote one i.e. many ransomewares come from the same author. So he could gain a reputation by signing messages saying that yes, this is our ransomware, we always unlock after receiving the payment.

>If it gets big enough then people just hear from each other if paying unlocks the data or not.

The idea would be to create "fake" ransomware that looks exactly like the real one

>The best way to end ransomware is to get serious about security

No matter how serious you get there always gonna be bugs, there isn't a single piece of mass distributed software in human history without bugs. That said, we should try to improve security of software but expecting it to be THE solution is wrong.

>Also, I imagine it gets easier after you wrote one i.e. many ransomewares come from the same author. So he could gain a reputation by signing messages saying that yes, this is our ransomware, we always unlock after receiving the payment.

Forging a signature is not that hard.

If forging digital signature is not that hard, then you can release some great scientific paper moving crypto decades ahead, or alternatively you can make billions.

I was talking about pixel-made signatures; you know, the one the user actually sees when the computer its already infected; not cryptography between public/private keys; otherwise its a chicken and egg problem; how do you know what signature its the "real"; you google it and hope nobody added its own search results? Go to the official website of the ransomware developer?

The ransomware can present the key fingerprint for example.

But even without it, there are so many options, e.g. timestamp signed message on the blockchain before the release. After just one confirmed message you don't care about pretenders because people can check if the signature matches with the previous message.

I think you are overestimating the technical capacities of the average randomware victim.

It's enough that some technical people verify it. The average random victim gets his info from news sites and more technical friends.

Is it not cryptographically possible to create a transparent provably-operational decryptor on top of something like ethereum?

It's a marketing issue. People likely to get hit with ransomware are incredibly unlikely to understand what that means. Hell, even main devs have trouble writing contracts, so even if a user knew there was a smart contract, verifying it would be another thing. So it'd get reduced to "guys on Twitter said this one works".

I like the idea though.

Since you can't store the private key needed to decrypt the files in ethereum, I can't think of how to do this.

All blockchain state is public, since it needs to be calculated by and verified by all nodes, so there's nowhere to stash a private key without revealing it.

It's a lot more likely that the attackers make money via the markets than via a direct ransom.

It's like saying best way to fight heroin addicts is to supply market with poisoned heroin. No heroin users - no problems!

The US government did this with alcohol during prohibition,

> by the time Prohibition ended in 1933, the federal poisoning program, by some estimates, had killed at least 10,000 people.

http://www.slate.com/articles/health_and_science/medical_exa...

A non-strawman analogy would be to sell fake heroin that looks exactly like heroin but does nothing at all. This analogy is more exact because the resource you are wasting its the same: money (not lives as in yours)

except that getting malware isn't addictive...

Your information is tho