During my presentation at the 45th Annual Heckerling Institute on Estate Planning on January 12, 2011, I mentioned that a strong password plus strong encryption can make it practically impossible to access an incapacitated or deceased person’s data if you don’t know the password. I gave a few examples of this during the presentation, and, in this posting, I want to provide more details to underscore the importance of planning ahead for passwords in the estate planning process.
Basically, encryption scrambles computer data so that you can’t read it without a password. Encryption can scramble a single data file or it can scramble an entire hard drive (or other storage media). With “weak” or “insecure” encryption, the data can be unscrambled relatively easily without knowing the password. Basically, a computer can guess all the possible passwords within a reasonable amount of time, which makes the encryption insecure. With “strong” encryption, it’s practically impossible to unscramble the data because it takes too long to guess all the possible password variations.
In the 1970s, 1980s, and 1990s, the U.S. Government used Data Encryption Standard (DES) to encrypt and protect data. DES uses a 56–bit key. Think of this “key” as the number of possible password variations. A 56–bit key has 256 possible password variations, which is about 72 quadrillion password variations (72,057,594,037,927,936 variations). In 1997, a $10,000 prize was offered to the first team that could unscramble a single DES encrypted message, and the winners were a team of scientists and about 78,000 volunteers who linked their computers together and unscrambled the message in about two months (link). By a 2008, special-purpose computer, built for just a few thousand dollars, could unscramble a DES–encrypted message in under a day (link).
Now, the U.S. Government uses the Advanced Encryption Standard (AES) for classified and top secret information, and you can use the same encryption on your home computer. AES uses 128–, 196–, or 256–bit keys, so there are significantly more possible password variations for AES than DES. For example, a 128–bit key has 2128 possible password variations, which is about 340 undecillion password variations (340,282,366,920,938,463,463,374,607,431,768,211,456 variations).
At the Heckerling Institute, I mentioned an example from a Technology Paper published by Seagate on 128–bit Versus 256–bit AES Encryption. Seagate makes hard drives that are protected by AES encryption. For their example, they assume there are 7 billion people, each person has 10 computers, and each computer could guess 1 billion passwords per second. If the password on a single 128–bit AES–encrypted file could be guessed after trying only 50% of the possible password variations, they conclude it would take 77 septillion years (that’s a 77 with 24 zeros after it) to guess the password.
However, I believe they made an error in their computations, and the result should be 77 billion years to the guess the password. They show their computations on page six of their Technology Paper. It looks like they divided by the number of seconds in a year (I’m using 31,556,926 for this number) instead of multiplying by the number of seconds in a year. You can tell because the number of “Total combos per second” (password guess attempts per second) is larger than the number of “Total combos per year” (password guess attempts per year). Clearly, there should be more password guess attempts made in a one-year period of time than in just one second. So, by multiplying by the number of seconds in a year instead of dividing by that number, it would take about 77 billion years to guess the password of a single 128–bit AES–encrypted file or hard drive (77,022,341,629.5 years).
According to NASA, the age of the universe is about 13.7 billion years, give or take 0.13 billion years. So, with the example above, it would take about 5.6 times longer than the age of the universe to guess the password of a single 128–bit AES–encrypted file or hard drive. If we want to guess the password of a second 128–bit AES–encrypted file or hard drive, it would take an additional 77 billion years! That’s why I say that a strong password plus strong encryption can make it practically impossible to access a person’s data if you don’t know the password.
On the other hand, using a “weak” password wipes out the protection of strong encryption. For example, if the password is one of the 470,000 or so entries in Webster’s Third New International Dictionary, Unabridged, or in The Oxford English Dictionary, Second Edition, or if it’s one of the 5,000 or so most commonly–used passwords or a variation on these, then even a typical home computer could probably guess the password of an AES–encrypted file in just minutes.
If we applied the same assumptions above (70 billion computers each guessing 1 billion passwords per second) on a single 56–bit DES–encrypted file or hard drive, it would not even take a full second to guess the password (0.0005 seconds)! Technically, the difference between 56–bit DES and 128–bit AES should be even larger than this because it takes more computational power to guess a single AES password than it does to guess a single DES password in general, so the same computer should be able to guess significantly more DES passwords per second. But, for our purposes, comparing less than a second to guess a 56–bit DES password versus 77 billion years to guess a 128–bit AES password is dramatic enough to illustrate the difference between “weak” versus “strong” encryption.
If 5.6 times longer than the age of the universe isn’t long enough for you (using 128–bit AES encryption), consider using 256–bit AES encryption. If we applied the same assumptions above (70 billion computers each guessing 1 billion passwords per second) on a single 256–bit AES–encrypted file or hard drive, it would take about 26 quindecillion years (26,209,344,715,487,043,370,350,318,887,362,908,146,482,868,240,272 years) to guess a single password!
To tie these examples to current events, consider WikiLeaks and its editor–in–chief Julian Assange. According to news reports by CNN and AlterNet in December of 2010, Mr. Assange recently released a 1.4 GB data file on the Internet scrambled with 256–bit AES encryption. It’s been called an insurance policy or “poison pill” in case he is arrested, assassinated, or his Web site is shut down. If that happens, his colleagues will release the password to unscramble the data. He’s hiding his secrets in plain sight, where anyone can try to guess his password. But, with the strong encryption method he selected (and presuming a strong password), he knows that it’s practically impossible with current technology to guess his password and access the protected data.
So, what should fiduciaries and family members do if a person becomes incapacitated or dies with encrypted data and did not plan ahead for passwords? Hire a computer expert to try to access and recover the data. This can be expensive and time–consuming. Also check for backups of the data on an external drive (or other storage media) or at an online backup service provider. The bottom–line is that, if the person used a strong password and strong encryption but didn’t plan ahead for his or her passwords, it’s probably not possible to access the data.