You are making that up. Of course there is such a mechanism. What do you think software signing is but a way for the good guys and only the good guys to run code on your device? It's exactly the same. All Apple has to do is make a piece of software that only runs on this phone and sign it. Then nobody can be attacked except this one phone. And it had to go through a warrant and all!

And to your last point: this software is not secure. That's why we have this backdoor option.

The problem with that thinking is that there is no firm definition of the "Good Guys", and the guys you think are good today, may not be so good tomorrow.

> this software is not secure. That's why we have this backdoor option.

It's not perfectly secure, but still awfully secure if even the US Government can't break it themselves.

> still awfully secure

I believe they were just sloppy:

http://abcnews.go.com/US/san-bernardino-shooters-apple-id-pa...

"San Bernardino Shooter's iCloud Password Changed While iPhone was in Government Possession"

The information wasn't "awfully" secured.

After watching this whole thing play out, I'm seriously wondering whether 'sloppy' was actually 'on purpose because this is a great chance to establish precedent'.

Now wait. If the passcode is complex then it is secure, which is a very typical requirement for security.

Which doesn't change regardless of the outcome of this FBI case.

Indeed, people who valued their security highly, like terrorists, would still have security, but the common person would not.

> All Apple has to do is make a piece of software that only runs on this phone and sign it.

(Emphasis added)

Just wondering how they can manage to produce something that will only run on this phone?

Easy. Since iOS software must be signed by Apple they could simply hard code a check for a specific device ID, and immediately exit if it doesn't match. Then they sign that specific binary.

That isn't really easy.

99% of the phone can be tampered with when running, after the signature is checked at install. They'd need to put that check in the secure enclave to make sure that it wasn't simply bypassed. That might not even be possible on that model.

By your definition then, the iPhone doesn't grant real security since Apple is capable of installing this modified iOS version.

So given that the iPhone is not secure, what do you think about whether Apple should execute this court order?

The iPhone does grant real security, it just requires the user to choose an encryption key that isn't trivially brute forced. (And depending on the circumstances, forego use of the fingerprint reader.)

If the dead attacker had a twelve character passphrase, any amount of help from Apple would be utterly useless. Apple could write all the back-doors they liked but if the content is encrypted and the key is unknown, brute forcing is the only option.

The addition of the secure enclave makes it potentially infeasible to extract the UID so that the brute forcing could be performed on a powerful supercomputer. It depends how well the enclave has been hardened against firmware changes. Can the enclave receive a seamless firmware update while in a locked state? Can the enclave firmware be rewritten to output the UID to the main CPU? We don't know.

The iPhone is secure when used with a complex passcode (something like 11 years to crack an 8 character password), even if Apple aids the FBI. So individuals who know and care enough, like terrorists, will still be just as secure.

But common people won't. So no, Apple should absolutely not destroy the security of hundreds of millions of individuals just because some government agency wants a shot at access to some dead guy's phone.

What happens when the PRC or Saudi equivalent of the FBI demands the same for an FBI agents phone? Why should the populace and Apple have to suffer for making the world more secure?

> So no, Apple should absolutely not destroy the security of hundreds of millions of individuals

Wait, how did you get to that point?