The leaks published by The Shadow Brokers (Wikipedia, New York Times) are the perfect example of why we should minimize the sensitive information kept by the government. I posted about this in the context of encryption backdoors two years ago. But whether we’re talking about encryption backdoors, security vulnerabilities or hacking tools, the most reliable way to avoid data leaks is to not store the data in the first place. It can’t be stolen if it doesn’t exist.
A simplistic way to do risk analysis for data leaks is to ask “what are the chances this will happen?” But that’s wrong. Data leaks will happen. It’s not a question of “if” it’s a question of “how often and how severe.”
We can look at the WannaCry ransomware attack for a real-world example. The Wikipedia article cites damage estimates of hundreds of millions of dollars. We don’t know how long the NSA has been building hacking tools, but I’d be surprised if they started before the year 2000. Are we ok with $100 million dollars in damages/lost productivity worldwide every 17 years?
What if it was every 10 years? Every 5 years? What if the damages were worse?
I strongly believe the potential harm from withholding security vulnerabilities far exceeds the potential gain from attempting to keep them private. The NSA shouldn’t be stockpiling security vulnerabilities or building hacking tools, and they shouldn’t have backdoors into encryption. The advantages do not justify the disadvantages.