In my opinion there are two major dangers of electronic commerce, fraud and the rape of civil liberties by companies capable of correlating information from multiple sources about a single person.
Fraud is possible only if the security of electronic commerce fails.
The destruction of civil liberties and misuse of information is only possible if companies can correlate information about individuals. These fears are compounded by attempts to make all encrypted documents observable by a single body like the SKIPJACK encryption system by the US National Security Agency 1, which gives the US National Security Agency the ability to decode all messages encrypted with their system.
The possible benefits of electronic commerce are the increased choice for the consumer, the possibility of improved security, and the possibility of standardization of payment methods. For a business it offers cost savings in transportation of physical value carrying tokens (coins!), simplification of payment methods, and a larger market.
Security:
There are two types of security commonly used in e-commerce, passwords/numbers, and encryption. To get money from an account a person has to have a number or password, generally it's a number, a credit card number for example. This information is supposed to be known only to the owning consumer, and temporarily to the organization the consumer is buying from. for example Credit Card information acts as a key to transfer money from an account. The other, very different, type of security is encryption. Encryption enables sensitive information to be sent publicly with knowing that only the body you are sending to will be able to decode the information.
Encryption:
Encryption Techniques have evolved greater in last 50 years. Encryption is the process of changing a message so that it can no longer be recognized without decrypting it. The decrypting process is hard without some sort of key, or clue, so that the the process of encryption allows you to control the flow of information.
The RSA crypto-system is becoming widely used, it has some very desirable properties, I won't give details of the maths, a good explanation can be found at: http://www.rsa.com/rsalabs/
Two other types of strong encryption are also available:
DES Encryption at http://www.rsa.com/des/
and elliptic curves at http://www.rsa.com/rsalabs/ecc/html/elliptic_curve.html
I'll concentrate on RSA encryption as it is more widely used and I understand it more clearly.
The encryption process involves the creation of two keys, called the public and the private key. The public key allows anyone to encrypt a message which is only decryptable with the private key. The RSA encryption rely's on the fact that it is difficult to factorize, large numbers. A second property of the RSA crypto system is Digital Signatures. A digital signature of a document is a way of encoding information about message using a private the private key, it is then possible for anyone with the public key to check a message to see if it was signed by the private key. An extension of the Digital Signature is called a "Blind Signature" this is a method where the data to be signed can be encoded further making it impossible for the signer to know exactly what it is they are signing.
Internet Secure Data Transfers
Currently security for transactions across the web tends to be in the form of a secure server.The protocol used is SSL (Secure Sockets Layer) which is an implementation of the RSA Crypto System (see http://www.rsa.com/ for details on the technical definition of rsa encryption). The server generates a private and public key, the public key is sent to the clients machine, this enables the client to encrypt the information being sent in such a way that only the server can decode it as it requires the private key to decode.
While the SSL is considered sufficiently secure to transfer credit card credentials, these systems are often not well constructed: while the transaction of the credit card details to the server is encoded, often the sending of the credit card holders details to the business from the server is not encoded2 . However it is generally considered, by net experts, safe to transfer credit card details across the net anyway as it's very unlikely that a hacker will intercept the data and obtain credit card details3 . Public opinion is more skeptical, from one internet software sales company4 suggest that one third of people ordering online don't trust the process.
About Electronic Commerce
Electronic Commerce has many forms, generally it is taken to mean either internet related commerce, including many virtual cash systems or Mondex style e-cash cards are also an example of electronic commerce in a different form. Both electronic-cash-cards and internet financial systems require security. Security for these systems takes the form of a method to stop 3rd party's from intercepting or spending someone else's money.
Currently, the classical cash system is still in use for most small commercial transactions, but it is now commonplace for large transactions to be done through some sort of electronic transfer.
The internet has grown exponentially over the last few years, making it an important area for electronic commerce. Many commercial transactions take place over the internet, these can be in the form of digital cash, virtual cash, credit card transactions, and incorporates nearly all aspects of electronic commerce. If fraud does occur it's difficult to prosecute any theft as it can occurs overseas where the laws are different. Unless active security is used it's a trivial task for a hacker to intercept a transaction and get credit card details. Serious social aspects for local businesses, and possible difficulty with competition.
There is a distinction that should be drawn between credit or debit schemes and token or cash systems 5, which rely on the exchange of some identifiable object, real or virtual, inherent given value. Mondex is an electronic cash system that aims to completly replaces the classical form of cash, but maintaining the usage structure. The Mondex card acts as the wallet and the cash all in one. A consumer withdraws money from the bank into the Mondex card, where he/she can then use the Mondex card as for purchases. Mondex is a token-based scheme developed by NatWest and uses encryption to stop fraudulent tampering of the cards. There was no security for the transaction itself on the Mondex card, however recently photo's have been placed on the card to stop anyone from using the a lost or stolen card. (see ".Net" June 1995 p34 to 60). Mondex offers a number of benefits for the party's involved. For the consumer it saves having to have the get change, and offers security advantages over cash, with a recent lock function being built in (although not in use at the University Of Edinburgh). For the business it saves money on having to have physically secure transportation on cash. Mondex also claim that it will speed up que's, so saving time. Mondex are also planning on pushing there influence onto the web, enabling people to buy goods using Mondex over the web.
While Mondex is supposed to be tamper proof there were roomers, which Mondex have not replied to my queries about, that in fact some card fraud has happened as recently as 1998. Because the Mondex card rely's on encryption for the tamper proof security if any advancements are made to code-cracking then it may leave all Mondex cards horribly insecure, causing large economic difficulties while standard are changed.
DigiCash, while also being a token-based system is quite different from Mondex, instead of relying on an existing currency it enables the owner of the software to create a new currency. While this could create breakthrough's in economic structure allowing the development of new financial systems it also poses a host of new currencies and standards. DigiCash is based on work by David Chaum, who in my opinion has proposed a good solution to the problem of the destruction of civil liberties by organizations capable of correlating hundreds of sources of data about a person building a manipulative structure6 .
Freedom Of Information?
One of the biggest and most controversial area's is privacy of information. The tendency is for organizations to collect and correlate data about people. The increase in electronic commerce and with the use of credit cards, given increasing amounts of information to companies, often more than people would like. It is the balance between privacy of information and the required public information that needs to be addressed.
David Chaum has proposed an interesting method of keeping privacy while still being able to authenticate payments. It is in my opinion a viable solution provided code breaking encryption techniques don't improve greatly.
A brief Explanation of the system proposed7 :
![[image]](pic1.gif)
This creates a bank note that can be checked easily for it's authenticity. However this on it's own would enable the bank to see every purchase that was made. a Solution that gives privacy to the consumer is has been developed by David Chaum, he calls it "Blind Signatures". It enables Alice to get the banks signature of the number without the bank knowing the number. By doing this privacy can be kept by the consumer.
![[image]](pic2.gif)
David Chaum also poses a refinement of this system where not every transaction must be verified by the bank. It involves some information being given about the identity of the spender when a note of currency is spent, where only if the same note is spent twice can the owner of the note be identified. This keeps privacy while preventing Fraud.
The Pitfalls
There are some concerns that are not addressed in the articles by David Chaum, namely what will happen if some method of breaking RSA cryptography efficiently is discovered. There is already two likely methods that may be able to do so in the foreseeable future: Quantum Computers8 and DNA Computing9 . DNA computing has already been shown demonstrated as a method which could crack codes in 1994. Quantum Computing is further away technologically, but similarly could be used to crack all popular encryption schemes in a matter of very quickly.
A Second concern I had with David Chaum's suggested system, assuming that neither Quantum Computer nor DNA Computing are successful in code cracking, is that if a group of hackers were to break the banks signature, in any way, they would be able to issue themselves with as much currency as they wished, causing the banks to have to change there private key making all currency redundant! The R5 encryption (using a 54 bit key) was cracked within two weeks by computers working together across the internet. This surly shows that much larger keys are needed for safe transactions for the next few years.
Electronic Commerce is growing and there's nothing that can be done about it easily. The dangers of using encryption in all commercial transactions is if a code-breaking breakthrough occurs then it will destruct the infrastructure developed. It also appears that there is a conflict of interests between some organizations and people's desire to keep information private. David Chaum's system appears to be a logical and intelligent system, providing that no code-breaking advances are made, it may well be able to strike the balance that will keep both organizations and people happy.
Bibliography
NewScientist 4 July 1998 p 23
NewScientist - 19 Sep 1998 - p23
David Chaum's Article in Scientific American, August 1992, p96
David Chaum on Electronic Commerce: How Much Do You Trust Big Brother?
email: internet-computing@computer.org
".Net" Magazine, Issue 7
".Net" Magazine, Issue 8
ICM Conference 1995, Peter Buck, analyst with Hyperion Systems.
RSA home page at: http://www.rsa.com/
Firewall FAQ at: http://www.tis.com/Home/Firewall's/FAW.html
Applied Cryptography, 2nd Edition, by Bruce Schneier, ISBN 0-471-11709-9
Foot Notes:
[1] See NewScientist 4 July 1998 p 23
[2] (this information is from an interview with Martin Davis manager of Matrix UK a software sales company which sells software aver the internet, contact Martin Davis by email: martin@astral.demon.co.uk).
[3] See ".Net" Magazine, Issue 7, p 54.
[4] Matrix UK. See footnote 1.
[5] See ICM Conference 1995, Peter Buck, analyst with Hyperion Systems.
[6] See ".Net" Magazine June 1995 p 59
[7] This was written from reading: Applied Cryptography, 2nd Edition, by Bruce Schneier, ISBN 0-471-11709-9, particularly Section 6.4; David Chaum's Article in Scientific American, August 1992, p96; and David Chaum on Electronic Commerce: How Much Do You Trust Big Brother? email: internet-computing@computer.org
[8] See NewScientist - 19 Sep 1998 - p23 and Applied Cryptography, 2nd Edition, by Bruce Schneier, ISBN 0-471-11709-9
[9] See Applied Cryptography, 2nd Edition, by Bruce Schneier, ISBN 0-471-11709-9