Dear Editors of Communications,

 Attached is a submission intended to be a "General Interest Article" for
your magazine. The manuscript is in PDF format. 

 The title of the manuscript is: 

	Telling Humans and Computers Apart (Automatically)
				Or
		  How Lazy Cryptographers do AI 

 The manuscript is about the ideas behind the technology that Yahoo is
using to prevent bots (or automated agents) from registering for accounts
in their system.  We introduce the concept of a CAPTCHA: a program that
can differentiate between humans and computers based only on interactions
over a computer network; a kind of Automated Turing Test. 

 Several websites have started using our technology since Yahoo
implemented it, and several others are using similar ideas to fight bots: 
Paypal, Altavista, etc.

 Bots, or automated agents have become a big problem in the web. We've
heard, for instance, that MSN is attacked by a bot that obtains 400,000
email accounts every day. Our technology is the solution to this problem. 

 Our research has sparked the interest of people from several areas within
computer science, including optical character recognition (OCR), computer
vision, synthesis and understanding of spoken language, cryptography,
network security, and even the hacker community. The first workshop on the
topic was held at Xerox PARC (now PARC) on January 2002, and another
workshop is being planned for January 2003, to be held at Carnegie Mellon 
University.
 
 Communications is the first magazine, journal or conference in which our
research will appear, and we believe it's the right place for it. The
ideas are provocative, yet easy to explain to the general CS audience.

 Below are the names of four experts that we consider appropriate to
review our submission. Since there is no field of Automated Turing Tests,
and because our research spans different areas within CS, these
individuals are experts in different fields: cryptography, computer
vision, natural language, and optical character recognition (OCR). In
addition, we include the name of the Chief Scientist of Yahoo, who has
worked in close collaboration with us on this project. 

Moni Naor (cryptography)
Professor
Department of Computer Science & Applied Math 
Weizmann Institute of Science                           
Rehovot 76100, Israel
naor@wisdom.weizmann.ac.il

Henry S. Baird (OCR)
Principal Scientist and manager of the Document Image Decoding Area,
Information Sciences & Technologies Lab, PARC, Xerox
3333 Coyote Hill Road 
Palo Alto, CA 94304 USA 
baird@parc.xerox.com

Roni Rosenfeld (natural language)
Associate Professor
School of Computer Science
Carnegie Mellon University
5000 Forbes Avenue
Pittsburgh, PA 15213 USA
roni@cs.cmu.edu

Jitendra Malik (vision)
Professor
Computer Science Division
University of California at Berkeley
725 Soda Hall
Berkeley, CA 94720-1776
malik@cs.berkeley.edu

Udi Manber
Chief Scientist
Yahoo! Inc.
udi@yahoo-inc.com

 A short message acknowledging the receipt of this submission would be
greatly appreciated.
	
 Sincerely,

		Luis von Ahn, Manuel Blum and John Langford

		Department of Computer Science
		Carnegie Mellon University		
		5000 Forbes Avenue
		Pittsburgh, PA 15213
		{biglou,mblum,jcl}@cs.cmu.edu