Discussion about this post

User's avatar
matt3298's avatar

Hoooooh boy. Hi Jeff. You know I have some opinions here and have been playing in the encryption issues in and out of govt for quite some time.

There is no real USG position on "strong" encryption or E2E as it gets vigorously debated internally due to the different mission needs. FBI wants access to evidence when they want it. They are very public about this and how (what they call) warrant proof encryption aids criminals. CISA tried to come out with a supportive statement for E2E but ended up watering it down to play the middle.

At NIST, the agency making encryption standards, the main issue is not to address this in the math, as you call it. The actual encryption itself must be trusted and, if there is a decision to do something for access, that should not be required in the algorithms. At that point, why use a US algorithm? Let use the Chinese standards instead and go to the larger market.

The IC wants everyone to just shut up. They end up getting what they need with folks who use bad, broken, outdated or incorrect implementations.

While crypo is your friend, it's actually a difficult technology to implement.

Hard crypto is good: Good crypto is hard.

Now, what about the talk of season two of Acolyte????

Side note on times they are a changing. It's not just AI taking up all the VC money and O2 in the room but we are about to do some significant encryption changes across the infrastructure in the near future on our crypto. On a positive note, we were very good at abstracting encryption away from users, so it just happens. On a negative note, we were very good at abstracting encryption away from users, so its hard for changes to just happen.

Expand full comment
Zoltan's avatar

Quite intriguingly, the people who are the most vocal advocates for backdoors in crypto are also the most particular about the one they use not having any.

Expand full comment
2 more comments...

No posts