Hello, welcome to the Wednesday, October 2nd, 2019 edition of the Sandstone Storms, Stormcast. My name is Johannes Ulrich, and I'm recording from Jacksonville, Florida.

Well, to start out with today, we got an interesting blog post by researchers at the Rural University of Boholm and they're looking into how to decrypt or

actually gain access to the decrypted content of encrypted PDF files.

Kind of notable is that they actually don't really decrypt the PDF.

So what they found was not an attack against the encryption algorithm really being used as

much as an attack against how PDFs are composed.

First attack is pretty straightforward.

The problem with encrypted PDFs that is being exploited

here is that PDFs actually may contain mixed encrypted and decrypted parts.

So what they're doing in this attack is they're adding some JavaScript that's not

encrypted to the PDF. Once the recipient decrypts the PDF,

the JavaScript is then used to exfiltrate the content. This is very similar to some of the

attacks that we have seen against encrypted emails, for example. Now, the second attack they came

up with is actually attacking a weakness in the

encryption, but again, the victim is actually decrypting the document for the attacker. And the

problem they're exploiting here is that PDF encryption uses the cipher block chaining encryption mode without any integrity checks.

If you're encrypting without actually checking the integrity of the message, then there's

always a possibility that the attacker will be able to modify the message and add additional

content.

Now to make this work, you do need some known plain

text in the PDF. Now, given the PDF structure and in more recent AS version 3 documents,

there's a permission structure at the beginning that can be used. It's not that terribly hard

...